CCIAP Project A-804

Impacts of Storms & Winds on Transportation in Southwestern Newfoundland

Norm Catto; Geography, Memorial University

Dale Foote, Dermot Kearney; Environment Canada

Wade Locke; Economics, Memorial University

Brad DeYoung; Physics & Physical Oceanography, Memorial University

Evan Edinger; Geography, Memorial University

Dan Ingram, Jeff Karn; Environmental Science, Memorial University

Jennifer Straatman; Geography, Memorial University

2006

1

CCIAP Project A-804

Impacts of Storms & Winds on Transportation in Southwestern Newfoundland

Table of Contents

Introduction …………………………………………………………………….. 1

Geomorphic Impacts of Waves and Storms on the Southwestern

Coastline of Newfoundland, including Port-aux-Basques Harbour ………… 12

Analysis of Tidal and Storm Surge record at Port aux Basques Harbour …. 28

Records and Physical Impact of Wreckhouse Events ……………………….. 54

Climate and Meteorological Analyses ………………………………………… 63

The Value of Economic Loss Associated with Weather-Related Delays

on the North Sydney-Port aux Basques Marine Transportation Link

in 2003 and 2004 ………………………………………………………... 112

Summary and Discussion ………………………………………………………. 117

References ………………………………………………………………………. 124

Appendix A: Station Histories ………………………………………………… 134

Appendix B: Seasonal Graphs ………………………………………………… 136

Appendix C: Wind Rose Diagrams ……………………………………………. 144

- 1 -

Impacts of Storms & Winds on Transportation in Southwestern Newfoundland

CCIAP Project A-804

Norm Catto; Geography, Memorial University

Dale Foote, Dermot Kearney; Environment Canada

Wade Locke; Economics, Memorial University

Brad DeYoung; Physics & Physical Oceanography, Memorial University

Evan Edinger; Geography, Memorial University

Dan Ingram, Jeff Karn; Environmental Science, Memorial University

Jennifer Straatman; Geography, Memorial University

1. Introduction

One of the key questions for transportation throughout Newfoundland is whether

storm and wind activity is increasing, in frequency or magnitude or both. Other locations

around the Atlantic Ocean have shown differing results, with some areas influenced by

increased storminess and others showing no change over the past 50 years, thus making

extrapolation to eastern Canada impossible.

All of the island of Newfoundland is dependent on maintaining regular Marine

Atlantic ferry service through the harbour at Channel-Port aux Basques. All road traffic

for the remainder of the island must pass along the Trans-Canada Highway from Port aux

Basques northeast to Corner Brook through the Wreckhouse area, noted for its strong

winds (“Wreckhouse Events”). At present, storm activity causes disruptions to ferry

operations at Port-aux-Basques, with ships compelled to remain at anchor outside the

harbour for two days or more, unable to dock until the wind subsides. During storm or

wind events, transport truck traffic is unable to travel along the Trans-Canada Highway

through the Wreckhouse area, as wind strength is sufficient to overturn large vehicles.

Increased storm severity and/or frequency would result in increased disruption to

both ferry and road traffic. Without a better understanding of the likelihood of increases

in either storm frequency or magnitude, Marine Atlantic and trucking operators are

unable to adapt effectively to the uncertain impact of climate variability. This uncertainty

ultimately affects all residents of Newfoundland who are dependent on this major land

2

and harbour transportation link. Transportation in this key region of southwest

Newfoundland is thus vulnerable to disruption from both storm surge and wave-induced

damage at the ferry terminal, and from wind events interfering with highway traffic.

1.1 Objectives

The objectives of this research directly related to marine storm events were:

• Assessment of the frequency and magnitude of extreme storm surges and high

tide events in southwest Newfoundland

• Documentation of decadal- to century-scale change in sea level, storm surges,

and extreme high tide events caused by climate change and variation

• Determination of the vulnerability of Channel-Port aux Basques, and the Marine

Atlantic ferry terminal in particular, to future extreme storm surges and high

tide events

The objectives directly related to highway transportation, as well as marine service,

were:

• Assessment of the frequency, magnitude, and wind direction of extreme wind

events in southwest Newfoundland

• Documentation of decadal- to century-scale change in extreme wind events

caused by climate change and variation

• Determination of the vulnerability of transportation infrastructure in southwest

Newfoundland to extreme wind events

Investigations to accomplish these objectives within CCIAP Project A-804 were

divided into the following components:

1. Co-ordination and community liaison (Catto)

2. Geomorphic Impacts of Waves and Storms on the Southwestern Coastline of

Newfoundland, including Port aux Basques Harbour (Catto, Ingram, Edinger)

3

3. Analysis of Tidal and Storm Surge record at Port aux Basques Harbour (Ingram,

DeYoung, Catto, Edinger)

4. Record and Impact of Wreckhouse Events (Catto, Straatman, Foote, Kearney)

5. Climate and Meteorological analysis (Foote and Kearney, with contributions as

acknowledged)

6. Economic impact analysis (Locke)

This report is organized to discuss each of these components. Chapter 1 introduces

the study region. Geomorphic impacts of waves and storms on the coastline, and in Port

aux Basques Harbour, are considered in Chapter 2. Analysis of the tidal and storm surge

activity at Port aux Basques Harbour is presented in Chapter 3. The record and physical

impacts of Wreckhouse events are discussed in Chapter 4. The climate and

meteorological data are analysed in Chapter 5. Economic impact analysis is presented in

Chapter 6. The results are summarized and discussed in Chapter 7.

1.2 Study Region

Investigation focused on the area of southwestern Newfoundland extending from

Channel-Port aux Basques to Wreckhouse (Fig. 1.1, 1.2). The communities of Channel

and Port-aux-Basques were amalgamated in 1945 to form a single community, officially

named ‘Channel-Port aux Basques’ but commonly referred to as Port aux Basques. In

2001, the population of Channel-Port aux Basques was 4,637, an 11.6% decline since

1996 (http://www12.statcan.ca/english/profil01). The declining population reflects a

general limitation of economic opportunity in southwestern Newfoundland, a trend which

has continued since 2001 (Government of Newfoundland and Labrador, 2005).

Port aux Basques has served as a harbour since the early 1500s. It was chosen as

the main ferry port and terminus for the Newfoundland Railway in 1898, and has served

as the main surface entry point to Newfoundland since that time. Construction of the

Trans-Canada Highway (completed in 1965) eventually rendered rail service redundant,

with passenger service ceasing in 1971 and freight service in 1988. Port aux Basques

4

Cape North

North Sydney

BurgeoChannel Port-aux-Basques

Cape Ray

Cabot Strait

Gulf of St. Lawrence

N

0 25 km

Figure 1-1: Satellite image of the Cabot Strait (NASA, 2002) and its location in relation to Atlantic Canada (inset; NRCan, 2005).

remains central to the Newfoundland economy, however, by virtue of its position at the

terminus of the Trans-Canada Highway in Newfoundland, and as the only port which has

year-round ferry service to the island.

Marine Atlantic operates a total of four ships between Port-aux-Basques and

North Sydney NS (Figure 1.1). Currently, three large ferries (M.V. Caribou, M.V. Joseph

and Clara Smallwood, and M.V. Leif Ericson) run continuously during the peak season

(generally from June to September) and two remain in service for the whole year. The

5

Figure 1-2: Aerial photograph of Port-aux-Basques Harbour (circa 1983). Note the ferry docked at the wharf (top left) and the orientation of the breakwater structures (photo from Government of Newfoundland and Labrador). M.V. Atlantic Freighter is dedicated to commercial traffic and freight, operating year-

round. From 2000 to 2004, the service transported an average of over 434,000

6

passengers annually (Marine Atlantic, 2005), and in 2005, an estimated 500,000 people

used Marine Atlantic’s ferry service through the port.

The Trans-Canada Highway extends northwestward from Port aux Basques,

paralleling the coast. The small coastal community of Cape Ray is accessed from the

highway, as is JT Cheeseman Provincial Park. Northwest of Cape Ray, the highway

turns northeast, passing through the Wreckhouse area flanked by Table Mountain, Sugar

Loaf, Cook Stone Hill, and Red Rocks Hill (Figure 1.3). The Wreckhouse area has been

noted for its extremely strong wind events since initial construction of the Newfoundland

Railway. Disruptions to rail service were recorded, with cars and engines forced off the

tracks, and highway traffic is periodically disrupted at present. The Trans-Canada

Highway leaves the Wreckhouse area south of the small community of St. Andrews.

The combined Port aux Basques – Wreckhouse route currently represents the major

year-round surface transportation link between mainland Canada and the island of

Newfoundland. It is the only link for highway traffic. Disruptions of this route, either by

wave and storm surge activity at Port aux Basques or by strong Wreckhouse wind events,

have resulted in severing the supply line to the remaining 500,000 residents of the island

of Newfoundland, from St. Andrews to St. Anthony and St. John’s. Study of the

potential for future interruptions resulting from weather events, climate variability, or

long-term climate change, is thus of importance for all residents of the island of

Newfoundland.

1.3 Present Regional Climate

Southwestern Newfoundland lies within the West Coast climate zone of the

province (Banfield 1981, 1993; Damman, 1983). The area is under the influence of a

cool humid mid-boreal climate (modified Köppen-Geiger Dfb), a consequence of the

influence of the Gulf of St. Lawrence and the dominant southwesterly winds.

Climate data is collected in Port-aux-Basques by an automated weather station

located approximately 1 km west of the harbour on Base Road, in the Mouse Island area

(47° 34' N, 59° 09' W, 39.70 m asl; Environment Canada, 2005). Daily mean

temperatures vary between -5°C in February and 15°C in August (Environment Canada,

7

N 0 100 200 metres

Figure 1-3: Nautical chart of Port-aux-Basques Harbour. Note the breakwater structures and the narrow shipping lane (Canadian Hydrographic Service, 1998).

8

2005; Catto, 2002a, 2002c; Catto et al., 2006). Although mean February sea surface

temperatures are less than 0°C along the entire coastline (Markham, 1980), development

of sea ice is restricted (Cote, 1989), allowing all shorelines in southwesternmost

Newfoundland to be affected by waves and storm surges throughout the winter (Catto,

1994a, 1994b, 2002a, 2002b; Ingram, 2004). The beaches of southwestern

Newfoundland thus are subject to reworking and modification throughout the entire year.

Mean annual precipitation along the coastline is 1500 mm, with 1570 mm recorded

at the Port aux Basques station (Environment Canada, 2005), of which 15-25% falls as

snow (343 cm snowfall at Port aux Basques). Strong winds and periodic thaws limit the

mean annual duration of snow cover to less than 70 days in coastal areas, although snow

cover in the interior persists for up to 5 months annually. Statistically, the driest month is

April (mean monthly precipitation 90-100 mm), and the wettest months are November

and December (mean 140-155 mm during each). The combination of substantial

precipitation and low summer temperatures, however, result in limited evaporation during

the summer months. The excess of precipitation over evaporation locally exceeds 100%

during some summers, as well as throughout the winter months, creating a perhumid

moisture regime and allowing wetlands to develop.

During the spring and summer, from March to August, the prevailing winds are

easterly. During the autumn and winter, from September to February, the prevailing

direction is from the west. The mean annual wind speed is 24.4 km/h (Environment

Canada, 2005) and higher magnitude wind events are common. According to the 30-year

Canadian Climate Normals, an average of 69.5 days per year experience winds exceeding

52 km/h and an annual average of 33.6 days experience winds in excess of 63 km/h

(Environment Canada, 2005). Easterly winds tend to have the greatest influence on ferry

traffic into Port-aux-Basques Harbour, and are also primarily responsible for triggering

Wreckhouse wind events.

Detailed discussion of the data related to potential patterns of climate variation

and change over the past 50 years is presented in Chapter 5 of this report.

9

1.4 Harbour Configuration and Port Infrastructure

Port-aux-Basques Harbour is open to the southeast, with an embayment extending

northeastward at its head, and is approximately 2 km2 in area (Figure 1.2, 1.3). The

mouth of the harbour is 900 m wide at the seaward limit. The southeastward orientation

makes the harbour susceptible to winter easterly storms. This vulnerability has been

mitigated with the construction of a breakwater, consisting of two separate sections,

confining the opening midway through the channel. At its narrowest point, the channel

opening is less than 400 metres wide (Figure 1.3). The effectiveness of the breakwater is

enhanced by the presence of two natural islands, Vardys Island situated in the centre of

the inner section of the harbour, and Pikes Island to which the breakwater is joined. The

two islands form important buffers to wave action.

The Marine Atlantic ferry terminal is located at the northwestern corner of the

harbour, on the former Ford’s Island. which has been connected to the mainland to make

room for the terminal infrastructure. To the east of the ferry terminal, and along the

eastern shore, the harbour is largely undeveloped. To the south of the terminal, extending

outwards in the harbour, the shoreline has been extensively developed. Numerous

wharves, including the Federal Government Wharf, Ring Bolt Cove Wharf, a recreational

board walk outfitted with a performance stage and concession stands, the Port-aux-

Basques Harbour Authority Wharf, and the local fish plant facilities are present.

Additionally, the outer area of the harbour is lined with private property with homes

constructed very close to the shoreline; some only 1 m above the high water line.

The harbour underwent major infrastructure development and improvement in the

1950s in preparation for larger ships and increased traffic. Along with the construction of

a new terminal building, the parking area was expanded and a new wharf and breakwater

were built. The L-shaped pier had a total length of 317 metres and enabled the

simultaneous docking of the southern Newfoundland coastal boats, servicing Burgeo and

ports east, and the larger Nova Scotia ferries (Windsor, 1956). The harbour was dredged

in 1953.

Construction of the breakwater was completed in July 1956. The structure was

assessed to have essentially eliminated undertow in the docking area, and overall docking

conditions at the pier were estimated to have improved by 40%. The breakwater extends

10

134 m across the harbour and is 21 m wide. At the end of the structure the water is more

than 10 m deep. The breakwater rises about 6 m above the waterline (Windsor, 1956).

Figure 1-4: Port-aux-Basques Harbour looking southeast. Note the breakwater extending across the harbour and a section of the Marine Atlantic dock in the foreground. Vardys Island is to the left and to the right is the Transport Canada vessel traffic tower and headquarters (photo taken June 2005).

Figure 1-5: The town of Channel Port-aux-Basques and the west side of the harbour. Note the ship (M.V. Atlantic Freighter) docking at the Marine Atlantic ferry terminal and the proximity of some of the homes to the ocean. (photo taken June 2005).

11

0 1000 m

N

Tide gauge

Climate station

Figure 1-6: Topographic map of the Port-aux-Basques area. The locations of the automated tide gauge and climate station are identified with yellow circles and labelled (Geomatics Canada, 1986).

12

2. Geomorphic Impacts of Waves and Storms on the Southwestern Coastline of

Newfoundland, including Port-aux-Basques Harbour

Norm Catto, Dan Ingram, Evan Edinger

2.1 Introduction: Storm surges and sea level rise

Newfoundland, including the southwest coast, has been subjected to many

significant storms, resulting in property destruction (Forbes et al., 2000; Catto et al.,

2003; Catto and Hickman, 2004; Catto, 2006b). Hurricane Juan (2003), the most

economically destructive hurricane event in Atlantic Canadian history, killed eight people

and was responsible for at least $200 million in damage to Nova Scotia and Prince

Edward Island (Environment Canada, 2004). Historical events, such as the Great

Hurricane of 1775 in eastern Newfoundland (Stevens and Staveley, 1991; Stevens, 1995;

Ruffman, 1995) provide ample evidence of the impact of extreme storms and storm

surges.

The effectiveness of any particular storm at a location depends upon the angle of

wave attack, the number of previous events during the season, and other local factors.

Adjacent beaches can exhibit very different responses to a particular storm, as was

evident on southwestern Newfoundland beaches impacted by Hurricanes Gustav (2002)

and Frances (2004). Beaches separated only by a headland varied widely in the amount of

geomorphic response to storms (e.g. Catto et al., 2003).

The northern Atlantic Ocean has been undergoing an increase in hurricane

frequency and magnitude since 1995 (Goldenberg et al., 1997, 2001; Landsea et al.,

1998; Debernard et al., 2002; Emanuel, 2005; Webster et al., 2005). However, the

relationship between changes in hurricane frequency and magnitude, and increases in air

temperature or sea surface temperature (SST), is not clear at present, and consensus does

not exist. Although causal links between SST changes and hurricane frequency and

strength have been suggested (e.g. Sugi et al., 2002; Trenberth et al., 2003; Knutsen and

Tuyela, 2004; Trenberth, 2005), other researchers have expressed reservations and

recognized uncertainties (e.g. Swail, 1997; Shapiro and Goldenberg, 1998; Pielke et al.,

2005; Webster et al., 2005). Regardless of the uncertainty of future changes in hurricane

13

frequency and magnitude in response to climate change, it is apparent that the North

Atlantic is currently undergoing a period of increased hurricane activity.

A storm surge is defined as the elevation of the water resulting from

meteorological effects on sea-level (Murty,1984; Pugh, 1987) The storm surge elevation

is the difference between the observed water level during the storm and the level that the

astronomical high tide would normally rise to in the absence of storm activity (Forbes et

al., 2004). The Port-aux-Basques area has experienced surges of a variety of magnitudes,

including several large ones, over the past 100 years. The return time of a surge that

exceeds 1 m is approximately 10 years, and every 50 years a surge will occur that

exceeds 1.5 to 2 m (Charles O’Reilly, personal communication). The full effect of a

surge may not be experienced unless it coincides with high tide, particular a high spring

tide. Storm surges are associated with both winter storms (e.g. January 2000) and

summer tropical storm systems (e.g. Hurricane Gustav, 2002).

Rising sea level allows waves and storm surges to penetrate further inland,

modifying beach morphology and sedimentology and resulting in both coastal erosion

and damage to infrastructure. The observed changes in sea level result from the interplay

of climate change, involving melting of the polar glaciers, thermal expansion of the

ocean, and consequent increase in the water volume in the oceans, with isostatic response

to past glaciation. With the exception of northernmost Labrador and Lake Melville,

Atlantic Canada is now subsiding, resulting in relative sea level rise, transgression and

flooding by the ocean.

Short-term (decades) changes in sea level can be assessed through study of tide

gauge records. At Port aux Basques, the tide gauge record from 1960 through 2005

indicates a mean rate of sea level rise of approximately 3.3 mm/a (Fig. ccc). This rate is

comparable to those observed at Charlottetown and Halifax (e.g. Shaw et al. 1998, 2001;

Bruce, 2002; McCulloch et al., 2002) and in eastern Newfoundland (e.g. Catto, 2006b,

2006c; Catto et al., 2003). Evidence of marine transgression is reflected by enhanced

erosion along many Atlantic Canadian beaches, including those to the northwest of Port

aux Basques (Catto, 2002) and in the Burgeo area (Ingram, 2004), and inundation of

terrestrial peat deposits and trees.

14

In low-lying areas with coastal development, erosion and flooding are

significantly increasing and will continue to increase (Daigle et al., 2006). Consideration

of the geological factors, rate of sea level rise, amounts of coastal erosion, wave climate,

and tidal regime allow calculation of the sensitivity to sea-level rise of shoreline

segments (e.g. Gornitz et al., 1993). This assessment has been completed for Atlantic

Canada as a whole on a broad regional scale (Shaw et al., 1998), and more detailed

assessments have been conducted for specific segments of the coastline (e.g. Chmura et

al., 2001; Catto et al., 2003; Daigle et al., 2006; Shaw, 2006). The observed sea level

rise in Atlantic Canada, and the projected rise for the future, depends upon the interaction

between the changing volume of the oceans and glacioisostatic activity.

2.2 Storm Surge Events in Southwestern Newfoundland

The coastline of southwestern Newfoundland has been subject to at least 5 storm

surge events since January 2000. The largest of these was the storm surge of 21-22

January 2000, associated with the same event that caused substantial damage in Prince

Edward Island and New Brunswick (Forbes et al., 2000; Bruce, 2002; McCulloch et al.,

2002; Parkes et al., 2006). Among others, lesser surge events resulted from Hurricane

Gustav (12 September 2002) and Hurricane Frances (2 September 2004).

The largest storm surge in recent years to affect the Port aux Basques region

occurred on 21-22 January 2000. Unfortunately, the precise magnitude of the surge

cannot be determined in Port aux Basques Harbour, as the tide gauge was out of service

at the time (see discussion in Chapter 3 and Ingram, 2005).

This event caused extensive damage and was the worst storm to hit the region

since 1974 (Bateman, 2000; Ryan 2000). The peak sustained wind velocity was recorded

at 96 km/h at 1:30 a.m. NST on 22 January, with a maximum gust of 137 km/h. During

this time, a wave-rider buoy recorded waves offshore to be approximately 15 m in height.

Environment Canada meteorologist John Macphee conducted a field assessment of the

Port aux Basques area following the event, and estimated that a rogue surge wave of

about 18 m height hit the community at around 3:00 a.m NST (Macphee, personal

communication; Bateman, 2000; Ryan, 2000).

15

The surge primarily impacted the Channel Head area of the town and caused the

greatest extent of damage. Homes received considerable damage, including flooding and

the removal of siding and roofing shingles. Vehicles were damaged by flying debris and

inundation, and exterior structures were destroyed. The 8 tonne Fisheries & Oceans

Canada mariner fog alarm buoy was shifted more than 1 km into the harbour. Coast

Guard equipment mounted on Channel Head received damage extending to 21 m asl,

suggesting that the wave impacted at ca. 16-18 m elevation, with salt spray extending a

further 3-5 m upward. Severe destruction to the walkways and other equipment at the

Coast Guard installation, however, at about 18 metres above sea level, was directly

caused by the wave (Bateman, 2000; Forbes et al., 2000; Ryan, 2000).

Overall, residential damage for this surge event alone totaled at least $280,000.

Damage to municipal infrastructure, primarily the sewage outfalls, was assessed at over

$400,000 (Bateman, 2000; Municipal and Provincial Affairs, 2000). The total economic

damage figure was estimated at more than $1 million (Len LeRiche, Public Safety and

Emergency Preparedness Canada, personal communication).

The actual monetary costs of disasters are always greater than the quoted values.

Reported figures are commonly are commonly the direct cost of the buildings, property,

and public infrastructure (The H. John Heinz III Centre, 2000; Kumar et al., 2001; Parson

et al., 2003). Discrepancies between insured values, personal expenses covered by

government compensation, and the actual or replacement value of the property lost can be

significant (Kreibich et al, 2004; Hickman, 2006). Monetary values of buildings and

infrastructure are readily determined, using assessed values for taxation or insurance.

Property located outside buildings, such as vehicles, is frequently not covered in total

value by government compensation programs or private insurance. Damage to exterior

property, uninsured items, and indirect costs are not included in the monetary values or

damage estimates quoted, especially those announced within a short time after the event.

The economic impacts of a disaster may continue for years after the event.

Damaged infrastructure may fail earlier than its designed lifespan (Kumar et al., 2001).

Prolonged health care and psychological counseling can be required for victims and

responders (Milne, 2002; Tapsell and Tunstall, 2003).

16

Hickman (2006) conducted an assessment of costs resulting from flood damage in

selected Newfoundland communities, and noted that indirect costs substantially outweigh

those resulting solely from damage to infrastructure. In June 2005, the cost of the

January 2003 flooding at Badger NL was estimated at $8.2 million (Paul Peddle, EMO

NL personal communication). This figure is more than 25 times the original estimate of

potential physical damage estimated by FENCO (1985), corrected for inflation, although

the 2003 flood waters reached almost exactly the elevation assumed in the 1985 study.

Much of the difference is represented by more accurate assessment of indirect costs not

associated with immediate response, rescue, and restoration of structures to pre-flood

conditions.

Hickman (2006) calculated that the March 2003 rain-on-snow flooding event in

Corner Brook, which resulted in $1.4 million in direct (recorded and announced)

damages, actually had an economic impact on the order of $5 million, although no deaths

or serious injuries resulted. Similar recorded damage totals resulted for the Burin

Peninsula from the impact of Hurricane Luis (1995) and for Torbay for Tropical Storm

Gabrielle (2001).

The actual costs of the flooding events studied in Newfoundland were more than

twice the reported figures in all instances. Applying a similar standard to Port aux

Basques would suggest that the actual financial impact of the 21-22 January 2000 storm

surge event would approach $ 3 million. This figure does not account for any losses

resulting in disruption to transportation between Port aux Basques and the remainder of

Newfoundland.

Social and community impacts of disasters are difficult to quantify. They are

commonly not recorded in estimates of ‘total cost’ of flood damage, which tend to be

expressed in monetary terms. Flooding hazards may impact individual perceptions of

personal security, producing stresses that are not directly expressed in monetary terms.

For community residents these impacts are as relevant as quantitative costs, and may be

more lasting.

The potential impact of a storm surge on Port aux Basques was the subject of an

exercise for municipal officials, first responders, emergency measures personnel, and

other responsible individuals on 13 June 2006 (see Seguin et al., 2006).

17

Prior to the January 2000 surge event, the most recent comparable surge occurred

on 20 October 1974. Logged water-level, atmospheric pressure, and wind data were

examined (Ingram, 2005). During this event, peak winds gusts reached the same

maximum as in the 2000 storm (137 km/h) and the maximum sustained winds were 84

km/h. Extensive damage was caused, including the destruction of the helicopter pad at

Channel Head. Sections of railway track along the coast at Osmond and north of Cape

Ray were damaged by waves and surge, which subsequently caused the derailment of two

diesel locomotives as they attempted to negotiate the damaged track (Bateman, 2000).

The lesser storm surge events which occurred in association with Hurricanes

Gustav (2002) and Frances (2004) primarily impacted the sand-dominated beach systems

at Grand Bay and JT Cheeseman Provincial Park, northwest of Port aux Basques. The

times of storm landfall on both occasions did not coincide with astronomical high tide,

reducing the storm surge heights. No damage was noted to installations in Port aux

Basques Harbour, although erosion of the dune-backed coastlines was notable following

Gustav.

The relative timing of storm surge events with respect to astronomical high tide is

critical. Hurricane Gustav did little damage in Charlottetown PEI, primarily because the

storm surge coincided with a lower tidal position, approximately 4 hours before

astronomical high tide. Environment Canada noted that if the storm surge had coincided

with high tide, water levels would have been comparable to those recorded during the

January 2000 event (http://www.atl.ec.gc.ca/weather/hurricane/gustav02_e.html), which

brought widespread flooding. Storm surges induced by hurricane activity thus are

potentially significant at Port aux Basques.

2.3 Tsunamis

Tsunamis are the products of seismic activity, and are not related to climate.

However, their potential impact on the southwest Newfoundland coastline is similar to

that for storm surges.

The 1929 Burin Tsunami killed 28 people in Newfoundland and 1 in Cape Breton,

a greater death toll than for any other Canadian seismic event (Ruffman, 1991, 1993,

1995a; Fine et al., 2005). The impact of the tsunami was felt along the south coast of

18

Newfoundland, from the Burin Peninsula west to Port aux Basques, although all the

Newfoundland deaths and almost all of the destruction occurred along the Placentia Bay

shore of the Burin Peninsula. Runup heights reached 13 m asl (Fine et al., 2005). Some

damage was noted to fishing vessels, stages, and gear in Burgeo, but there were no

injuries. The occurrence of other tsunami events, notably in 1864 at St. Shotts (Avalon

Peninsula) has been suggested (Ruffman 1995a, 1995b), but these have not been verified,

either by historical accounts or through geological studies.

Preliminary investigations at Burgeo have not revealed the distinctive tsunami

laid sands (TLS) that commonly mark landfall areas for tsunamis, both on the Burin

Peninsula (e.g. Ruffman, 1993; McCuaig and Bell, 2005) and elsewhere (e.g. Dawson,

1999). However, TLS deposits are generally preserved in isolated lagoons, salt-water

marshes, and other low-energy environments not influenced by repetitive wave erosion.

The lack of suitable preservation areas at both Burgeo and Port aux Basques Harbour

means that TLS deposits would not be expected to be preserved, even in areas that were

known to have experienced tsunami activity.

The occurrence of the Burin Tsunami demonstrates that tsunami waves, of

comparable elevation to the largest waves associated with storm surges, could strike the

coastline in the Port aux Basques and Burgeo areas. The total damage to property on the

Burin Peninsula was estimated at about $1 million in 1929 dollars (Liverman et al.,

2001), equivalent to ca. $11 million in 2006. This figure, however, must be regarded as

an extremely low estimate. In addition to the difficulties involved in estimating the actual

costs of damage from natural hazards, as alluded to above, the infrastructure destroyed

along the Burin Peninsula in 1929 consisted largely of wooden stages, small fishing

vessels, and houses. A comparable event today would result in much greater monetary

loss.

2.4 Coastal Erosion

The sensitivity of a coastline to erosion is governed by the energy and frequency

of wave action, and by the physiography and geological composition of the coastal zone.

The majority of the southwestern coast of Newfoundland lies within the Newfoundland

Uplands of Appalachia (Bostock, 1970). Eastward from Port-aux-Basques to Burgeo,

19

fjard morphology characterizes the shorelines developed on Palaeozoic granite and

associated igneous rocks (Riley, 1959; Dickson et al., 1996), with irregular embayments,

numerous skerries and small islands. Embayments are variable in depth, with irregular

profiles influenced by differential pre-glacial weathering and glacial scouring. This area

is generally resistant to coastal erosion, except where accumulations of sand backed by

coastal dunes are present, as at Sandbanks Provincial Park (Catto 2002, 2006a; Ingram

2004). Burgeo Harbour, flanked by exposed granite, is not subject to coastal erosion,

even during storm surges.

Metamorphosed Palaeozoic units (Port-aux-Basques schist, van Staal et al., 1996)

underlie the coastline extending northwestward from Port-aux-Basques to J.T.

Cheeseman Provincial Park and Cape Ray Cove. The structural weaknesses and

lithologic divisions in the metamorphic units generally trend NE-SW. Erosion, both by

coastal and terrestrial processes, has exploited the planes of weakness, creating a

shoreline marked by shallow coves separated by rock spurs aligned parallel to the

geological structure. This alignment mirrors that of the prevailing winds of the area,

resulting in topographic funneling of waves into embayments.

In areas where bedrock is exposed, such as Port aux Basques Harbour, weathering

is confined to terrestrial processes, primarily frost action. The coastline is not subject to

erosion by waves, even those resulting from hurricanes and storm surges.

Erosion has affected the dune-backed coastlines of Grand Bay West and JT

Cheeseman Provincial Park, northwest of Port aux Basques, and Sandbanks Provincial

Park, near Burgeo. Along the Gulf of St. Lawrence coast of Newfoundland, the

combination of rising sea level, increased human utilization of the coast for tourism

purposes, and limited offshore winter ice conditions have resulted in accelerated erosion

and degradation of the dunes and coastline (Pittman and Catto, 2001; Catto, 2002a,

2006b; Ingram 2004). Ingram (2004) noted erosion at the rate of 0.7 m per month

between December 2003 and April 2004 at Sandbanks Provincial Park, concentrated in

the late summer-autumn period, and in early winter before freezing of the beaches

occurred, providing some measure of protection. This rate of erosion represents a

substantial acceleration from previous estimates (Catto, 1994).

20

Figure 2.1: Coastal Erosion at Sandbanks Provincial Park, Burgeo. View looking to the northwest at the mid-section of the beach, 9 April 2004 (Dan Ingram)

Figure 2.2 Sand-dominated beach and backing dune, JT Cheeseman Provincial Park, showing mobilization of coarse clasts, evidence of recent storm-surge erosion.

21

The impacts of climate change and variation, storm events, and human activity on

J.T. Cheeseman and Sandbanks Provincial Parks, and the impact on Piping Plover

(Charadrius melodus) have been discussed in detail elsewhere (Catto, 2002a, 2002b). A

summary of the principal findings of these previous investigations is presented here.

In southwestern Newfoundland, coastal dune development and associated sandy

beach evolution is related to destabilization of littoral areas initiated by marine

transgression (Catto, 1994; Catto, 2002a). Consequently, rising sea level, ca. 3.3 mm/a,

is acting to destabilize the coast, increasing vulnerability to coastal erosion and storm

surges. The beaches of southwestern Newfoundland are subject to reworking and

modification throughout most or all of the year. The coastline typically remains entirely

ice-free until February, and during some years, it may remain ice-free throughout the

winter.

Removal of overlying fine sand has exposed underlying hard-packed medium

grained sand, and storm-transported shell hash, coarse sand, granules, and fine pebbles

now overlie many areas. Storm-driven waves have resulted in bank erosion in some

areas of the dune complex at JT Cheeseman in excess of 1 m between summer 1999 and

June 2002 (Catto, 2002a, 2002b). In addition to changes in grain size distribution

resulting from fluctuating longshore drift, storms, snow cover, aeolian activity, and

humans also cause changes in the sediments exposed on the beach.

Regional changes in temperature and precipitation are only secondarily

responsible for dune modification and erosion. Boreal conditions prevailed from mid- to

latest Holocene in southwestern Newfoundland (see Davis et al., 1987; Davis, 1993;

Anderson and Macpherson, 1994; Macpherson, 1995). The sedimentary structures within

the dune sequences throughout southwestern Newfoundland indicate that formative

processes remained similar throughout the period of dune accumulation. In the absence of

detailed chronological investigations, there is no evidence to suggest any substantive or

consistent change in formative processes over time. Local climate conditions governing

sediment flux, and human disturbance of the dunes, are more significant than are broad

changes in regional climate.

22

The available information for southwestern Newfoundland suggests that any

summer warming trend may be offset by increasing summer precipitation, and the overall

perhumid moisture regime is not expected to vary significantly under current climate

change scenarios (Lewis, 1997; Lines et al., 2003). Future conditions would not be

optimal for sand dune formation or expansion. Limited sand remobilisation from dunes

to coast over a long period will lead to gradual coarsening of the beaches. If beaches are

deprived of sand from longshore transport, there is no ready source of sand from

stabilized dune fields, and consequently the coarsening process will be accelerated under

storm wave reworking. As dunes and beaches are linked parts of a system, a reduction in

dune mobility results in a decrease of sand availability for beach maintenance. This

could have an impact on tourist utilization and perception of the beach systems (Catto,

2004).

The high winds associated with storms are accompanied by precipitation, which

simultaneously diminishes their effectiveness in eroding the higher dunes. Storm winds

are most effective at eroding previously existing trough and linear blowouts, and are

relatively ineffectual at eroding windward surfaces. In undisturbed areas, the strongest

storm winds tend to sweep across the dune surface, and may result in accretion where

small saucer blowouts are infilled by grainfall deposits upon quiescence.

The primary effect of storm activity on the dunes is to destabilize the dune

surfaces, rendering them vulnerable to anthropogenic disturbance. By removing

vegetation and sweeping debris from linear blowout bases, the storm winds create freshly

exposed surfaces that are subsequently available for human reworking. The propagation

of linear blowouts is facilitated by the strong storm winds, thus widening and deepening

the tracks and making them more attractive for human foot traffic. The storm winds also

rework and deepen deflation hollows. Deflation hollows influenced by anthropogenic

bonfires and quarrying are particularly susceptible to deepening and widening during

storms.

Natural impacts are related primarily to storm winds, waves, and surges. The

decrease in sediment influx from the dunes also plays a role, and the slowly rising sea

level has a lesser influence. These processes have resulted in erosion, coarsening of

23

sediment along the shoreface, and a gradual steepening of the beach profile. The natural

effects are currently muted by the generally low energy of the beach environment, and

the presence of sufficient time between storms in most years, with the exceptions of

1995, summer 1999-January 2000, and the period autumn 2004-autumn 2005, including

Hurricane Frances.

2.5 Sensitivity to Sea Level Rise

Assessment of the sensitivity of a shoreline to erosion resulting from sea level rise

involves consideration of several variables. Study of shorelines in the eastern United

States by Gornitz (1993), Gornitz et al. (1991, 1993), and of Canada by Shaw et al.

(1998) led to the identification of parameters which can be used to assess the sensitivity

of a shoreline to erosion. Shaw et al. (1998) list seven critical parameters:

• relief,

• rock and/or sediment type exposed along the shore,

• landform type (e.g. cliff, beach, salt marsh),

• tendency of sea-level change (amount of rise or fall per 100 years),

• shoreline displacement (laterally, expressed in m / a);

• tidal range, and

• mean annual maximum significant wave height

Shaw et al. (1998) assigned each parameter an equal weight, and ranked variations within

each from 1 (very low sensitivity) to 5 (very high sensitivity). By combining the scores

for each parameter, sensitivity indices (SI) can be calculated as:

SI = √ (product of scores of all 7 parameters / 7)

Thus, a shore with the least sensitivity to coastal erosion would have a SI of √ (1/7), or ~

0.38, whereas the greatest value possible is √ (5 x 5 x 5 x 5 x 5 x 5 x 5 / 7), or ~108.

24

Shaw et al. (1998) divided the coastline of Canada into three categories of SI.

Coastlines with low sensitivity had SI values of ≤ 4.9; moderately sensitive coastlines had

values between 5.0 and 14.9; and highly sensitive coastlines had values in excess of 15.0.

A single sensitivity index was calculated for all 2899 of the 1:50,000 map areas along the

Canadian coastline.

Throughout the analysis, Shaw et al. (1998) carefully indicate that the regional

nature of this investigation may serve to partially conceal local problem areas.

Investigation on a regional scale (c.f. Catto et al., 2003; Daigle et al., 2006) allows

further subdivision of parameters, assessment of their relative importance locally, and

designation of more specific areas for categorization.

Shorelines with high relief above sea level are relatively insensitive to erosion. In

contrast, shorelines with relief less than the mean significant wave height are susceptible

to periodic inundation, increasing the potential for erosion. Offshore of eastern

Newfoundland, the mean annual significant wave height is estimated at 7 m - 8 m (Neu

1982; see also Lewis and Moran, 1984), with the 10-year and 100-year values estimated

at 11 m and 15 m respectively. These values can be compared with the recorded 15 m

height at the offshore wave-rider buoy and 18 m inundation level associated with the

January 2000 event. Estimates of significant wave heights based on modeling tend to

underpredict extreme storm wave heights (Bacon and Carter, 1991; Cardone and Swail,

1995; Cardone et al., 1995), and wave heights in excess of 30 m have been recorded

offshore of the south coast of Newfoundland during storm events (Swail, 1997).

In this study, shorelines with relief less than 7 m are considered to have a very

high risk relief factor (5). Shorelines with relief less than 15 m are considered to have a

high risk relief factor (4), and those with relief less than 20 m are assessed as moderate

(3). Areas with relief of 30 m or less are assessed as having low risk (2), and those above

30 m are at very low or negligible risk (1).

The primary difference related to rock and sediment type along the coastline

distinguishes areas with relatively resistant bedrock (such as Port aux Basques Harbour)

from areas of coastal sand dunes (such as Grand Bay West and JT Cheeseman Provincial

Park). The resistant granite of the Burgeo area has a very low risk factor (1). The schist

surrounding Port aux Basques Harbour is less resistant to frost wedging, due to its

25

jointing and fracture pattern, but its sheltered position with respect to most storm tracks

limits its exposure, reducing the sensitivity (2). In contrast, areas with exposed aeolian

sand dunes flanking sandy beaches, open to the direction of prevailing winds, are very

sensitive to erosion (with a sensitivity factor of 5).

Landform type also influences the sensitivity to coastal erosion. The steep fjard

coastlines between Burgeo and Port aux Basques (1) and the rock platforms surrounding

Port aux Basques Harbour (2) are less sensitive to erosion than are the dune-backed

coastlines (4).

The tendency of sea-level change (amount of rise or fall per 100 years) also

controls coastal erosion. The rankings of sensitivity indices for these factors follow those

of Shaw et al. (1998). For Port aux Basques, sea level is currently rising at 3.3 mm/a, ca.

33 cm per 100 years. Coastlines with sea level rising between 21 and 40 cm/100a are

assigned a sensitivity factor of 4 by Shaw et al. (1998).

Sensitivity factors for shoreline displacement vary from very low (1) for accreting

shorelines, through low (2) for shorelines showing no net displacement, to very high (5)

for shorelines receding at rates in excess of 1 m/a. Values throughout the study region

for this parameter thus range from very high at Sandbanks Provincial Park (rates >> 1

m/a, Ingram 2004) to very low. At Port aux Basques Harbour, the bedrock shoreline has

shown no perceptible change, and small gravel fringes have accumulated in some

locations. The harbour is assigned a sensitivity factor of very low (1) for this parameter.

Tidal ranges in eastern Newfoundland lie within the microtidal and very lowest

mesotidal limits. Port-aux-Basques Harbour has a mixed, mainly semi-diurnal tidal

regime. The range during mean tide is 1.1 m, rising to 1.6 m during spring tide

(Canadian Hydrographic Service, 2004 and 2005). Similar microtidal regimes were

assigned a sensitivity factor of low (2) by Shaw et al. (1998).

The mean annual maximum significant wave height is also important in

assessment of the sensitivity index. Along the eastern Newfoundland coastline, the

significant wave height of 7 m - 8 m (Neu, 1982) establishes this parameter as having a

very high sensitivity ranking.

26

By combining the values for all seven sensitivity factors, an overall coastal

sensitivity index can be calculated. For Port aux Basques Harbour, the Coastal

Sensitivity Index is:

CSI = √ (3 x 2 x 2 x 4 x 1 x 2 x 5) /7 = 8.3

This value lies in the lower range of the ‘moderate’ sensitivity classification. For Port

aux Basques, the sensitivity factors relating to sea level history and wave activity are far

higher than the geological and topographical factors. As all factors are equally weighted

in the formula, the overall result suggests low-moderate sensitivity. Observations of the

ineffectiveness of storm surges and waves on erosion of this coastline, however, suggest

that the geological factors are locally dominant. Overall, Port aux Basques Harbour is

not sensitive to coastal erosion.

For Sandbanks Provincial Park, the Coastal Sensitivity Index is:

CSI = √ (5 x 5 x 4 x 4 x 5 x 2 x 5) / 7 = 53.5

This value indicates a coastline highly sensitive to erosion, which is in agreement with

observations at the park.

27

2.7 Summary

• Sea level is currently rising at Port aux Basques at ca. 3.3 mm/a.

• Rising sea level will allow successive storm surges to rise higher and penetrate

further inland.

• The coastline of southwestern Newfoundland has been subject to at least 5 storm

surge events since January 2000. The largest of these was the storm surge of 21-

22 January 2000.

• The estimated economic damage from the January 2000 event exceeded $1

million. However, study of other flooding events in Newfoundland indicates that

actual costs were more than twice the reported figures in all instances. Applying

a similar standard to Port aux Basques would suggest that the actual financial

impact of the 21-22 January 2000 storm surge event would approach $ 3 million.

This figure does not account for any losses resulting in disruption to

transportation between Port aux Basques and the remainder of Newfoundland.

• The shoreline at Port aux Basques Harbour is not sensitive to coastal erosion

resulting from sea level rise. Coastal erosion is occurring along dune-backed

sandy shorelines, such as those at JT Cheeseman and Sandbanks Provincial Parks.

28

3. Analysis of Tidal and Storm Surge record at Port aux Basques Harbour

Dan Ingram, Brad DeYoung, Norm Catto, Evan Edinger

3.1 Introduction and Methodology

Port aux Basques is prone to frequent storm activity, particularly intense winds.

Storm surges are commonly a consequence of such events. This chapter summarizes the

investigations of the long term impacts on water level height, assessing the role of tidal

variation on storm surge elevation. Detailed discussion is presented by Ingram and

DeYoung (2005) and Ingram (2005). The aims of this investigation were:

• to determine whether the tidal regime is changing over time;

• to determine the astronomical variations effecting any changes;

• to relate any residual change in tidal regime to changes in sea level and/or wind or

storm surge activity; and

• to assess the implications, particularly with respect to the impact on port and ferry

operations.

Statistical analysis utilized tide and climate data obtained from the Marine

Environmental Data Service (MEDS) of Fisheries and Oceans Canada, and Environment

Canada, respectively. The tide data dates back to 1935 but contains many large gaps,

particularly during the 1930s through to the 1950s. The set, however, is largely complete

from 1959 to present, although gaps exist due to mechanical problems with the gauge.

For the purposes of this study, records from the 1950s through to the end of April 2005

were analysed. Unfortunately, there is no climate data for the Port aux Basques station

prior to 1966.

The tide and pressure data was analysed using MATLAB (Version 6.5.1.199709

Release 13) in conjunction with T_TIDE. These programs were accessed through a

computer work station in the Department of Physics and Physical Oceanography at

Memorial University. T _TIDE was obtained from Professor Rich Pawlowicz, Ocean

Dynamics Laboratory, University of British Columbia (http://www2.ocgy.ubc.ca/~rich/).

29

3.2 Analysis

Initially, data was reconfigured for compatibility with MATLAB and T_TIDE.

This included converting the dates to the Julian calendar and the times from

Newfoundland Standard Time (NST) to Greenwich Mean Time (GMT). The data was

placed in annual increments for analysis using T_TIDE. Classical tidal harmonic analysis

(Godin, 1972) was completed on the raw, unaltered water level data. This separated the

predicted harmonic tide from the actual recorded water level, and thus separated the tidal

residual from the actual tide.

The dominant tidal constituents were identified from the predicted harmonic tide.

The frequency, amplitude, and phase of each constituent were separated and listed in an

ASCII file. Generally, 67 constituents were identified for each data set but the number

varied depending on the completeness of the record for any particular year.

For this study, only the most significant constituents (those with an amplitude

exceeding 0.0400 m) were investigated further. Eight such constituents were identified;

S2, M2, N2, K1, O1, K2, SSA, and SA. The frequencies of each are listed below in Table

3.1.

30

Table 3.1: Tidal harmonic description, period, and frequency of selected tidal

constituents (modified from LeBlond and Mysak, 1978).

Constituent

Tidal

Harmonic

Description

Period

(semi-diurnal,

diurnal, or longer)

Period

(mean solar

hours)

Frequency

(cycles per hour)

S2 Principal solar Semi-diurnal 12.00 0.0833333

M2 Principal lunar Semi-diurnal 12.42 0.0805114

N2 Longer lunar

elliptic

Semi-diurnal 12.66 0.0789992

K2 Luni-solar

declinational

Semi-diurnal 11.97 0.0835615

K1 Soli-lunar

declinational

Diurnal 23.93 0.0417807

O1 Main lunar Diurnal 25.82 0.0387307

SSA Solar semi-annual Longer period 4393.3

(182.6 days)

0.0002282

SA Solar annual Longer period 8767.6

(365.2 days)

0.0001141

An annual mean time series was plotted for the amplitude of each constituent. The

most evident trends or cycles were shown by K1, N2, M2, O1, and S2.

Once the predicted harmonic tide and its constituents were identified and removed

the raw residual water level data was analysed. It is this component that is of greatest

interest since it includes the atmospheric pressure component, amongst other factors, and

any associated storm surge events.

3.2.1 Residual Water Level

The residual, non-tidal data was zeroed by subtracting the mean from each data

point. Each annual time series was then prepared for filtering in T_TIDE, to remove any

non-tidal noise from the data. For filtering, each series had to be relatively complete,

31

generally with record gaps not exceeding one month. Years with large continuous gaps,

or numerous smaller ones, were omitted for the purposes of filtering and further analysis.

Filtering in MATLAB requires that any time series must be complete with no gaps, and

therefore, any gaps in the record had to be infilled with data that was interpolated. This

could be reasonably done with gaps less than one month in duration, but not with large or

frequent ones. For the purposes of this analysis, simple linear interpolation was used.

This did not fill the gaps with accurate water level data but rather simply filled the voids

allowing a filter to pass over without complications, as filters do not respond well to

discontinuity.

The Butterworth digital lowpass IIR filter (Etter, 1993) was used for each annual

time series. Butterworth filters have maximally flat passbands and stopbands and are

written as:

[B,A] = butter(N,Wn)

with N representing the order of the filter and Wn being the cutoff frequency normalized

to the Nyquist frequency (Etter, 1993). For the filtering of the water level residual, an

order of 6 (N) and a cutoff frequency of 0.03 cph, or roughly 33 hours, were used.

Although the filtered data removes other noise (such as instrumental error and

high frequency variability), permitting a more accurate statistical assessment of trends,

the raw water level residual is still utilised, particularly when dealing with storm and

surge events since the main interest is with the actual water level is those situations.

Therefore, the annual raw residual time series were used to determine annual maximum

events and the length of time that water levels exceeded particular thresholds.

3.2.2 Extreme Water Level Events

Two water level thresholds were used. The most important threshold is a

residual, or surge, that exceeds 60 cm. This is the widely recognised definition of a storm

surge event (c.f. Murty, 1984; Forbes et al., 2004). Figure 3.1 shows the number of

events that exceeded 60 cm each year and the number of hours exceeding this level.

The same data was also plotted using a threshold of 40 cm (Figure 3.2). This

provides an indication of the trend of water level events greater than the mean. Although

32

there may be no evident trend for actual surge events (those exceeding 60 cm), there may

be a large increase in higher water level events. Evident increases in high water levels

may not be sufficiently extreme enough to be classified as surges. Utilising a lower

arbitrary water level threshold in conjunction with the standard storm surge definition

allows better assessment.

Years with a complete data set are identified by the solid blue bars, and those

years with partial records are represented by the grey bars. Any years with significant

gaps, particularly during the winter months, have been omitted and marked with an

asterisk. A data set was considered complete (solid blue) if no gaps exceeding 24 hours

were present in the record, meaning that the yearly data set must be at least 97.3%

complete. The plots were excluded totally (marked by an asterisk) if the winter record

was missing more than 30 days in total, or if the annual data set was missing more than a

total of 90 days. Thus, the winter data set had to be more than 66.7% complete, and the

overall yearly record had to be more than 75.3% complete. All other years were

classified as having partial records (grey shading).

33

1960 1965 1970 1975 1980 1985 1990 1995 2000 20050

1

2

3

4

5

6

7

Year

Num

ber

of

Events

1960 1965 1970 1975 1980 1985 1990 1995 2000 20050

5

10

15

20

25

Year

Dura

tion o

f E

vents

(H

ours

)

(A)

(B)

* * * * * * * * * *

* * * * * * * * * *

34

Figure 3.1: The number of storm surge events (water level exceeding 60 cm) (A) and the number of hours the water level exceeds 60 cm (B) by year in Port-aux-Basques (1959 – 2004). The blue bars indicate a complete data set (less than 1 day total gap). Grey bars indicate partial data sets (a total gap exceeding 1 day but less than 90 days annually and/or 30 days during the winter). Asterisks (*) indicate omitted data sets due to very large gaps (more than 90 days total and/or more than 30 days during the winter).

35

1960 1965 1970 1975 1980 1985 1990 1995 2000 20050

5

10

15

20

25

Year

Num

ber

of

Events

1960 1965 1970 1975 1980 1985 1990 1995 2000 20050

20

40

60

80

100

120

Year

Dura

tion o

f E

vents

(H

ours

)(A)

(B)

* * * * * * * * * *

* * * * * * * * * *

36

Figure 3.2: The number of water level events exceeding 40 cm (A) and the number of hours that the water level exceeds 40 cm (B) by year in Port-aux-Basques (1959 – 2004). The blue bars indicate a complete data set (less than 1 day total gap). Grey bars indicate partial data sets (a total gap exceeding 1 day but less than 90 days annually and/or 30 days during the winter). Asterisks (*) indicate omitted data sets due to very large gaps (more than 90 days total and/or more than a 30 days during the winter).

37

3.2.3 Climate Factors

Atmospheric pressure data was converted to a water height equivalent (in metres)

for comparison with water level data. This allows an assessment as to how much, if any,

of the water level fluctuation is pressure dependent.

Wind speeds were converted to m/s and was used to calculate the drag coefficient

(CD) using the methods of Large and Pond (1981). The drag coefficient was reduced to

10 m height and neutral atmospheric conditions (CDN) (Large and Pond, 1981).

Once the CDN was defined, the zonal (u) and meridional (v) components of wind

speed and the wind stress was resolved into its zonal (τx) and meridional (τy) components:

τx = ρA·CDN·u(u2 + v2)1/2

τy = ρA·CDN·v(u2 + v2)1/2,

where ρA is the density of air (approximately 1.26 kg/m3).

Using the zonal and meridional components, the absolute mean wind stress (τ)

values could be calculated (in N·m-2 or Pa) as:

|τ| = (τx2 + τy

2)1/2.

As an absolute value the wind stress was plotted as a time series for composition

with the residual water level and atmospheric pressure data plots. For this study, only the

absolute values were considered. If specific storm events and associated impacts were to

be studied, it would be valuable to consider each directional component separately. This

would be particularly true when investigating the relationship between any given wind

direction and the recorded water level in Port aux Basques Harbour.

3.3 Interpretation

3.3.1 Tidal Cycles

The most obvious cyclic patterns are seen with the amplitudes of M2, O1, and K1,

all of which display two clear oscillations with a period of about 18 years each (figure

3.3, 3.4. 3.5). These are thus likely the dominant components of the 18.6-year tidal cycle.

A longer time series, however, would be required to make any definite interpretations.

38

1930 1940 1950 1960 1970 1980 1990 2000 20100.41

0.415

0.42

0.425

0.43

0.435

0.44

0.445

0.45

0.455

0.46

YEAR

Am

plit

ude (

m)

Figure 3.3: Amplitude for tidal constituent M2 (1935 - 2005) based on analysis of hourly data from each year.

39

1930 1940 1950 1960 1970 1980 1990 2000 20100.07

0.075

0.08

0.085

0.09

0.095

0.1

0.105

0.11

0.115

YEAR

Am

plit

ude (

m)

1930 1940 1950 1960 1970 1980 1990 2000 20100.065

0.07

0.075

0.08

0.085

0.09

0.095

YEAR

Am

plit

ude (

m)

Figure 3.4: Amplitude for tidal constituent O1 (1935 - 2005) based on analysis of hourly data from each year.

Figure 3.5: Amplitude for tidal constituent K1 (1935 - 2005) based on analysis of hourly data from each year.

40

For each constituent, the 18-year cycle can be attributed to the previously

identified tidal cycle, although the 2 separate 18-year oscillation sets are of different

variance. A longer time series that illustrates more of the cycles would be required to

make any interpretations. The same can be said for downward trend, as it cannot be

accurately explained without a longer time series.

3.3.2 North Atlantic Oscillation

The cyclic pattern of the constituents, particularly M2 that has the dominant

influence due to its relatively large amplitude, was also compared to the North Atlantic

Oscillation (NAO; Hurrell et al., 2003).

The M2 and NAO display a similar cyclic pattern with similar periods (Figure

3.6). The time series are also plotted along with the annual maximum surge. The annual

maximum surge bar plot has been colour coded on the basis of completeness of the

record. Years with no significant record (1964 and 1979) have been omitted completely.

To evaluate the relationship between M2 and the hours of exceedence of a 40 cm water

level and the NAO, scatter plots were constructed (Figure 3.7). From these plots it is

clear that there is no real correlation, although all three variables exhibit some cyclical

pattern.

41

1960 1965 1970 1975 1980 1985 1990 1995 2000 20050.41

0.42

0.43

0.44

0.45

0.46

0.47

M2 Amplitude

Wate

r Level (m

)

Year

1960 1965 1970 1975 1980 1985 1990 1995 2000 20050

0.2

0.4

0.6

0.8

1.0

Annual Maximum Surge

Year

Wate

r Level (m

)

1960 1965 1970 1975 1980 1985 1990 1995 2000 2005-6

-4

-2

0

2

4

6

NAO Index

Year

(A)

(B)

(C)

42

Figure 3.6: Yearly comparison of the amplitude of tidal constituent M2 (A), the NAO Index (B), and annual maximum surge height (de-tided residual) (C). Complete data sets and those with gaps are shown in blue and grey, respectively.

43

-6 -4 -2 0 2 4 60.4

0.5

0.6

0.7

0.8

0.9

1.0A

nnual M

axim

um

Surg

e (

m)

NAO Index

-6 -4 -2 0 2 4 60

20

40

60

80

100

120

140

NAO Index

Dura

tion (

Hours

)(A)

(B)

44

Figure 3.7: Scatter plots of the NAO Index versus the annual maximum surge height (A) and the number of hours per year that the water level exceeds 40 cm (B) (1959 – 2004). Note the lack of correlation between the two variables in each plot.

45

Scatter plots were also constructed comparing the NAO Index versus the number

of events exceeding 40 cm and 60 cm and the number of hours that the water level

exceeded 60 cm. No relationship was evident in either plot. Similarly, no relationship

was present when the NAO was plotted against the annual maximum surge either. These

results complement the work of Karn (2004) who indicated that the local significance of

the NAO is very variable and no clear relationship or trend exists with respect to wind

speed in Atlantic Canada.

3.3.3 Residual Water Level

The residual water level displays a very clear seasonal trend, as is evident when

the monthly mean residual water level is plotted (Figure 3.8). The fall and winter months

have a higher mean water level and a higher variance than do the spring and summer.

This is expected as the stormiest months are during the fall and winter, which would

cause increased surge activity and generally higher water levels. Based on mean seasonal

data (1959-2005), the ratio of the standard deviations of the mean winter (January,

February, March) and summer (July, August, September) water level is approximately

1.6, a considerable difference.

46

JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC-0.1

-0.08

-0.06

-0.04

-0.02

0

0.02

0.04

0.06

0.08

0.1

Month

Wate

r Level (m

)

Figure 3.8 Monthly mean water level (de-tided residual) and variance from 1959 – 2005

There is a clear seasonality shown by the data, with higher water levels in the fall

and winter months (Figure 3.8). Therefore, the plot displaying the mean annual water

level (Figure 3.9) should be used primarily to further emphasize the large variability in

the water level, as well as the challenges of working with an incomplete, very fragmented

data set.

47

JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC-0.1

-0.08

-0.06

-0.04

-0.02

0

0.02

0.04

0.06

0.08

0.1

Month

Atm

ospheric P

ressure

(H

eig

ht equiv

ale

nt in

metres)

Figure 3.9: Monthly mean atmospheric pressure (height equivalent in metres) and variance from 1959 – 2005.

3.3.4 Influence of Atmospheric Pressure and Wind on Water Level

Residual water level is greatly influenced by atmospheric pressure and local wind,

amongst other less significant factors. For Port aux Basques, its location with respect to

northerly and north-easterly tracking low pressure systems and their associated winds

provide many examples of this interaction.

Every storm and associated surge is unique. There is no one template or set of

guidelines outlining how a particular system will impact any given area. There is,

however, some strong relationships that exist between water levels and atmospheric

pressure and wind (Table 3.2). Atmospheric pressure, similarly to water level, shows

48

obvious seasonality. The mean monthly pressure, plotted as height equivalent, is

basically the inverse of the water data for the same period (1959-2005).

Table 3.2. Recorded and calculated data for selected storm events.

Recorded / Calculated Variable JAN 2000 OCT 1974 FEB 1969

Maximum actual (m) nd 2.15 2.54

Maximum residual (m) nd 0.51335 0.82421 Water Level

High tide prediction (m) 0.53636 0.49543 0.49543

Atmospheric Minimum recorded value (kPa) 96.98 97.73 98.24

Pressure Minimum height equivalent (m) -0.41787 -0.34918 -0.3044

Proportion of surge due to pressure (%) nd 68% 40%

Maximum speed (km/h) 96 84 93

Direction (deg) 180 170 110

Direction (descriptor) S S E Wind

Max stress (Pa) 1.9842 0.87116 1.1858

Seasonal sea-level rise is illustrated in the monthly mean water level, or is at least

an active factor, amongst others such as wind stress. The opposite is true during the

summer months. The variance is much less and the mean is much smaller. The obvious

seasonal difference in variance, for both the water level and atmospheric pressure, is

likely due primarily to the climatic regime. Port aux Basques is prone to fall and winter

storms and such isolated events will create anomalies in the data and thus increases the

variance. Conversely, the summer months are relatively calm and the isolated storms are

typically less severe resulting in less variability in the data.

The influence of atmospheric pressure is evident for specific storm events as well.

A relationship clearly exists as the water level and wind peak at relatively the same time

as the pressure dips to a minimum. This example displays a somewhat idealistic

representation of the relationship. Other factors, such as wind direction and lag times

attributed to low pressure system movement, are also involved. In this example, the

maximum surge of 0.824 m peaked at approximately 1530 hrs GMT on 4 February. This

practically coincided with the high tide, producing a combined water level of 2.54 m

above chart datum. This is the ‘worse case scenario’ with respect to destructive impacts

for storm surges. If the maximum surge occurred during low tide the actual water level

would be much less since the water level resulting from the harmonic tide would not be

49

factored in. When a high tide and surge coincide there is essentially a stacking effect and

the already elevated water level, as a result of the high tide, is further elevated due to the

surge. The impact is enhanced even further in the event that the high tide is the higher of

the two of the daily cycle and/or it occurs during spring tide conditions.

High winds were prevalent for about 24 hours during the 4 February storm, both

preceding and following the maximum surge. The peak sustained winds (93 km/h) were

logged at 0530 hrs GMT and exhibited a stress of about 1.19 Pa. These winds subsided

for some time later that day but returned again to reach a maximum of 89 km/h (a wind

stress of 0.93 Pa) at 1230 hrs GMT. Thus, the maximum surge lagged the peak winds by

about 3 hours. At this time the stress was gauged to be about 0.93 Pa. Peak winds,

however, primarily serve as a gauge of maximum intensity and the duration of significant

50

30 31 32 33 34 35 36 37 38 39-1

-0.5

0

0.5

1

Wate

r Level (m

)

Year Day

30 31 32 33 34 35 36 37 38 39-1

-0.5

0

0.5

1

Pre

ssure

Heig

ht

(m)

Year Day

30 31 32 33 34 35 36 37 38 390

0.5

1

1.5

2

Year Day

Win

d S

tress (

Pa)

0.824 m

98.24 kPa-0.304 m

89 km/h0.93 Pa

93 km/h1.19 Pa

(A)

(B)

(C)

51

Figure 3.10: Time-series plot of residual water level (A), atmospheric pressure (height equivalent) (B), and wind stress (C): 31 January – 9 February 1969. winds must be noted as well. Often the winds have weakened by the time of the surge

maximum but the sea state still demonstrates the effects of the previously intense winds.

As a result, it cannot be assumed that wind stress does not have a significant impact

simply due to a decrease in velocity at the time of the surge. Similarly to the impact of

atmospheric pressure, the effects on the water level often lag in time. The ocean does not

react instantaneously to changes in environmental or meteorological influences. It takes

time to adjust and thus water level values commonly, but not always, lag wind and

pressure changes.

3.3.5 Extreme Water Level Events

Extreme water level events are relatively common in Port aux Basques. During

the partially complete water level record, there were a total of 69 surge events (residual

water level greater than 60 cm) that persisted for approximately 230 hours over the 45

year period (1959 – 2004), an average of 1.5 events and 5.2 hours of surge conditions

annually. This estimate is less than the actual number, as many events were not logged

due to the very fragmented record. However, the extreme water level record for Port-

aux-Basques displays an oscillating pattern. Some years with a complete data record

indicate no extreme events, while others with only a limited number of water level

records show a large number.

3.4 Recommendations and Suggestions for Future Work

The methods and techniques used for this research were very applicable and

proved to be effective, but the data set was a limiting factor. With a relatively short and

fragmented data set, long term trends could often not be identified. This was particularly

true for changes with respect to storm surge frequency; either exceeding the 40 cm or 60

cm threshold, and maximum storm surge height. It was clear, however, that storm surges

are very variable over the duration of the data.

To make a better assessment of the long term trends, a long time series would

have to be analysed. Unfortunately, the tide gauge in Port-aux-Basques was not installed

52

until 1935 and the record is largely incomplete until 1959 and contains large gaps for the

duration of the record. Therefore, this project can possibly serve as the starting point, or

base-line, for continuing research as more data is collected at Port-aux-Basques. Ideally,

the time-series would be expanded at the end of each year. This, however, will mean that

an obvious trend cannot be established for many more years. The outcome is unknown,

especially as there are predictions that climate change is accelerating and its impacts will

likely be more apparent.

Another option would be to reconstruct the past record. This would involve

significantly more statistical analysis, but is possible. Any such re-construction would be

based on tide gauge records from other Atlantic Canada locations that were operational at

the time. The recent water level data from Port-aux-Basques would first be compared to

the data from other stations. Once a correlation is established the past gaps in the record

would be able to be roughly interpolated. If the other gauge records pre-date 1935, the

records preceding the first Port-aux-Basques data could be interpolated as well. Such

interpolations would not be perfectly accurate for event- or storm-specific water levels,

but could be used to determine the frequency and approximate magnitude of surges.

Thus, it will enable the input of more data to better establish a trend. Actual water levels

could not be determined due to the large local-variability in Atlantic Canada. The same

storm system often has a very different impact in different locations, even if they are

geographically close.

Future research could also be enhanced by upgrading the tidal network. The

gauges should be maintained in a manner that large gaps are not present and any gaps are

not as frequent in the record. This is especially true for the winter months when storms

are more prevalent and having a record of the water level is more critical. Essentially, the

present state of the Atlantic Canadian tide gauge network is not acceptable for scientific

research nor for public safety. This can be exemplified very clearly during the very

intense and destructive storm of January 2000. During this severe storm the gauge was

not operational, had not been for 130 days prior to the surge (since 14 September 1999)

and did not log data for another 38 days after. The total non-operational period for this

gap was 168 days, and many other gaps in the record exceed this. Notably, at least 3

other intense storms (wind stress > 1.5 Pa) impacted the area during this period, and these

53

water levels were not logged. The networks require adequate maintenance and associated

funding, in order for significant and productive research to continue and for public safety.

3.4 Summary

• There is no clear net change in tidal regime for the period of tidal records for Port

aux Basques (1935-present). There is, however, an indication of the de-tided

residual increasing over time (approximately 10 cm over the 46-year record; 1959

to 2005). Unfortunately, this increase cannot be accurately interpreted since the

data set is highly fragmented, but some of the increasing trend is likely due to

relative sea level rise in the area.

• There is a clear seasonal trend in mean water level. The mean and variance is

higher during the fall and winter than in the spring and summer.

• The atmospheric pressure exhibits a clear seasonal trend. The pressure is lower

with a higher variability during the late fall and winter.

• There is no obvious correlation between the NAO and the tidal constituents, such

as M2, or storm surge activity (neither by frequency in terms of number of events

or duration nor by maximum surge height).

• The annual number of events that exceed 60 cm shows no real trend and there has

been little change in the annual maximum storm surge elevation.

• The maximum water level for extreme events, such as the January 2000 storm,

that have been excluded from the data record can be roughly estimated using

basic statistical correlations based on the recorded meteorological conditions

along with the records of other previous storm/surge events.

• Tidal gauge networks require adequate maintenance and associated funding, in

order for significant and productive research to continue and for public safety.

54

4. Records and Physical Impact of Wreckhouse Events

Norm Catto, Jennifer Straatman, Dale Foote, Dermot Kearney

4.1 Previous Research

Wreckhouse Events have impacted surface transportation through southern

Newfoundland since the construction of the Newfoundland Railway, and numerous

anecdotes exist (Morris, 1977; Harding, 1992; Lingard, 1997). Studies and archival

documentation concerning these events, however, are surprisingly limited. Four previous

studies have considered the impacts on transportation and hydroelectric transmission in

the area.

Sutherland et al. (1963) investigated extreme wind events and train delays in the

Wreckhouse-St. Andrews area for the period 1956-1962, at the request of Canadian

National Railways. The wind data used in the study was from the weather station

observations at St. Andrew’s, which the authors related to weather conditions at

Wreckhouse. The meteorological component of the study incorporated the conclusions

of earlier research by Janz, using data obtained between 1948 to1958. Further discussion

of the relationship between conditions at Wreckhouse and wind data from surrounding

stations is presented in chapter 5 of this report.

The majority of the events occurred with southeasterly winds ahead of warm fronts

or trowels, and could develop quickly. The phenomenon was a result of the cold air

becoming trapped between the advancing warm front and the high terrain a few miles

inland. With continued pressure from the warm air, the cold air was forced through the

natural funnel at much greater than normal velocity. It was proposed that a mechanism

similar to the Bernoulli affect could be the cause of the strong winds.

The majority of the abnormal winds originated from the east or southeast, ahead of

warm fronts, and diminished rapidly with the passage of the front. They occurred

predominately from the late fall to early spring with few or no events during the summer.

Wind speeds increased very rapidly over short periods of time, reaching 144 km/h (90

mph) with gusts to 190 km/h (120 mph). A total of 130 gale force events were

documented between 1956 and 1962, of which 84% originated from the east or southeast.

These winds were responsible for the delays in rail traffic.

55

Sutherland et al. (1963) stated that the abnormal winds producing Wreckhouse

Events were assisted by funneling along the flanks of the Red Rock Hills, Sugar Loaf Hill

and Table Mountain. The presence of the Horizontal Wave Forest at Red Rock Hills was

associated with the wind regime. The climatology of Wreckhouse events is discussed

further in chapter 5 of this report.

Sutherland et al. (1963) were primarily concerned with delays to rail traffic (Tables

4.1, 4.2, 4.3). Days with wind speeds in excess of 65 km/h (40 mph) were compared with

traffic delay data obtained from the office of the Superintendent of the Line of the CNR.

Train delays occurred on 44 days with gale force (or stronger) winds recorded at St.

Andrew’s, and on 13 days with lower wind speeds. All of the train delays occurred when

the wind was from the east, southeast, or south. Most of the gales occurred during the

months of November through March. Over the 8 year period, the 57 delays represent an

average of 7.1 per year.

As traffic was delayed on approximately 33 % of the number of days with gale force

or stronger winds, Sutherland et al. (1963) also looked at the number of consecutive

hours the winds were strong, hypothesizing that perhaps the strong winds did not last

very long, and therefore did not conflict with the train schedule. They concluded that

there was insufficient data to explain why train delays only occurred on some of the days

with strong winds, but not others, and why delays sometimes occurred when the winds

were not as strong (i.e. less than 65 km/h). The longest train delay was 19.5 hours, the

shortest 1.5 hours.

Table 4.1 Number of train delays by month and year in the Wreckhouse area,

1956 to 1962

Month

Year Jan Feb Mar Apr MayJun Jul Aug Sep Oct Nov Dec Total

1956 0 0 0 0 0 0 0 0 1 1 2 1 5

1957 1 3 2 1 0 0 0 0 0 0 1 2 10

1958 1 3 0 2 0 1 0 0 0 0 1 1 9

1959 2 0 3 0 0 0 0 0 0 0 2 0 7

1960 2 5 0 0 0 0 0 0 0 0 2 0 9

1961 3 0 1 0 2 0 0 0 0 0 0 0 6

1962 0 1 1 1 0 0 0 0 0 0 3 4 10

Total 9 12 7 4 2 1 0 0 1 1 11 8 56

56

Table 4.2 Days of strong winds in relation to the number of train delays 1956 to

1962

Month Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total

Days of strong winds 23 16 15 10 9 3 0 0 4 9 18 23 130

Days of delays with

winds over 40 MPH 6 9 6 3 2 0 0 0 1 2 7 8 44

Days of delays with

winds under 40 MPH 3 3 1 1 0 1 0 0 0 0 2 2 13

Table 4.3 Trains delays in comparison with wind direction 1956 to 1962

Wind Direction N NE E SE S SW W NW Total

Days of strong winds 0 0 27 82 9 7 2 3 130

Days of delays with

winds 40 MPH or over 0 0 14 27 5 0 0 0 46

Days of delays with

winds less than 40 MPH 0 3 2 7 1 0 0 0 13 Weather Engineering Corporation of Canada (1982) examined the cause of the

extremely high winds at St. Andrew’s and elsewhere in the Codroy Valley that damaged

Newfoundland and Labrador Hydro transmission towers along TL-214, with a goal of

assessing design and maintenance. Using the data obtained from the St. Andrew’s

weather station for the period 1953-1966, the frequency distribution and extreme values

of the wind speed were determined as well as return periods of extreme wind events and

the probability of occurrence of any wind speed.

Rawinsonde data from Argentia, obtained during severe storms, was used to

calculate wind speed frequency distributions for 305 m (1000 ft) elevation and compared

with the surface observations at St. Andrew’s. The results indicated the wind speeds at St.

Andrew’s were stronger at a given return period than were winds 305 m above the

ground, in contrast to normal conditions elsewhere.

This study downplayed the role of funneling or channeling of the winds through

gaps in the high terrain as the cause of the extreme winds. Using wind data from other

recording equipment, the study indicated areas of divergent flow off the Long Range

Mountains were also regions of severe winds. It was proposed these winds were a type of

57

fall-wind caused by the rapid downslope flow off the Long Range Mountains. When east

or southeast winds off the interior plateau of the Long Range Mountains reached the

escarpment, would drop-off quickly to the low lying areas at Wreckhouse and St.

Andrews. Local channeling of the wind through river valleys modified the winds but it

was essentially a downslope wind phenomena. The study investigated the relationship

between downslope wind storms in the Codroy Valley and the energy of the air stream.

Theoretical relationships were applied for two storms and the results correlated well with

observed winds and damage to transmission lines in the Codroy Valley. The study also

investigated the extreme gustiness of the downslope winds and concluded the Boyd gust

formula fitted the observations at St. Andrew’s to a good approximation.

The primary objective of this study was to assess the design of Transmission Line

214. Consideration of the data indicated that TL 214 should be designed to withstand an

extreme sustained wind speed of 177 km/h, with gusts to 235 km/h.

TRO Engineering (2002) conducted a further assessment of TL-214 for

Newfoundland and Labrador Hydro. Based primarily on the Weather Engineering

Corporation (1982) report, it was recommended that TL-214 be redesigned to withstand

gusts to 235 km/h, rather than the original design standard of 177 km/h for individual

gusts. TRO (2002) conducted a design analysis based on the higher gust value, focusing

on phase spacing, horizontal insulator swing, and conductor design. A central concern

was ensuring that conductors had sufficient weight so that they would not be displaced by

strong winds towards the cross-arms, as contact between the two would result in design

failure.

ADI Limited (1994) conducted a study of Wreckhouse winds and their effect on

truck traffic, to assess the cost effectiveness of a wind warning system. Warning signs

and markers have subsequently been installed along the highway. The intent of this study

was to assess the magnitude of accidents and delays, and to discuss remedial measures,

rather than to analyze the meteorological situation.

Over the five year period 1989-1993, ADI Limited (1994) documented 17

rollover/high wind blowoff accidents. Of these, only 9 (59%) occurred during the winter

months of November through March, suggesting that high winds are possible during any

season. The estimated average cost of each accident, based on insurance claims for

58

vehicles and cargo, was ca. $50,000 in 1993 (estimated by AJ Bell and Grant Ltd.,

Insurance Brokers, Halifax), equivalent to ca. $70,000 in 2006. The mean annual cost

estimate from 1989-1993 was $170,000, equivalent to $236,000 in 2006. These figures

do not include uninsured losses, and also refer only to the value of the hardware involved.

The cost of delays was also estimated by ADI Limited (1994). At Doyles, high

winds force westbound truckers to remain parked until the winds subside. The estimate

of 25 such events per year, each one lasting approximately 6 hours, affecting

approximately 100 trucks per event, produces a value of 15,000 truckers hours per year

lost in delays in each direction, for a total of 30,000 hours lost in delays per year. With

adjustment for inflation between 1993 and 2006, the total estimated loss by truck traffic

delay each year is ca. $1.9 million. This figure does not include delays to smaller, non-

commercial vehicles, or costs involved with response to accidents or road maintenance

and repair.

4.2 Archival Research – Western Star

The previous studies involved relatively short investigations of specific forms of

transportation and infrastructure. Although estimates for highway traffic delays for the

period 1989-1993 are broadly valid for 2006, given the meteorological data available (see

Chapter 5) and the similarity in driving practices and vehicle design, comparison of rail

and road delays would not produce a valid estimate of the amount of change resulting

from changed climatic factors. Consequently, efforts were made to investigate archival

sources, in an attempt to determine if accident and delay frequencies had changed

throughout the period of railroad operation. Any such changes could then be considered

in light of changes in railroad equipment and maintenance protocols.

The Western Star newspaper (based in Corner Brook) was the most constant

source of data concerning derailments and delays. Publication of the Western Star started

in April 1900, and has continued until the present. Due to time constraints, and the lack

of detailed archival cataloguing of the newspaper, it was decided to investigate three

sample periods of two years each. This involved studying all of the issues from 1901-

1902, 1931-1932, and 1951-1952. All references to train delays due to any climate or

weather-related cause for the line between Stephenville and port aux Basques were noted.

59

These dates were chosen because they avoided the World Wars (1914-1918,

1939-1945) which would have limited the type of information published in a newspaper.

The earlier volumes of the paper had a reasonable account of such local information, but

later volumes, particularly those from the 1950s, had expanded to include more

international news, advertisements, and current clothing fashions, at the expense of local

news.

The data from 1901-1902 indicated a total of 9 recorded delays. This figure

represents the lowest possible estimate, as only the most extensive delays would be

considered newsworthy. The most significant event occurred on 25-27 November 1901,

and involved storm surge damage as well as high winds:

Friday 29 November 1901 Vol. 2 No. 62 [p. 3] “TERRIFIC GALE Held up

Express Train AT CODROY! Monday’s express to connect with the S.S.

Glencoe was held up at Little River station all night. A terrific gale swept

through the Codroy valley and carried the waters of the ponds in clouds of

spray for hundreds of yards. It seemed as though rain was descending in

torrents whereas really the night was comparatively clear. It was hardly safe

to be moving from car to car on the train, the force of the wind being so

great as to almost lift a man off his feet. In view of last year’s accidents, in

which one train was blown from the track and burnt up and another entirely

derailed, the railway people are acting prudently in stalling the trains during

such peculiar weather conditions ...”

[p. 4] “TIDAL WAVE Visits West Coast. Great Damage to Railway. Not

for seventeen years has the West Coast been visited by such a tide as that

experienced on Wednesday forenoon. Fortunately the day was calm and

there was no swell on, otherwise stores and dwellings would have been

floating about … at Grand Bay, or, to be more accurate, at Barachois, about

five miles east of Port au Basques, the terminus of the railway, the wind

blew with great force and the trestle work and gravel bed succumbed to the

ceaseless lashing of the waves. The ballast was washed from under the ties

and piles for nearly a hundred yards and all traffic became suspended. The

60

express for St. John’s with the Glencoe’s mails and passengers was stalled at

Port au Basques and did not leave until 5 o’clock yesterday afternoon,

reaching here shortly after four o’clock this morning.

“The express from St. John’s to connect with the Glencoe was held up at

Little River Wednesday night and was unable to reach P.A.B. The

passengers and mails were transferred to a train on the opposite side of the

washout and conveyed to Port au Basques, where they arrived at 5 p.m.

yesterday. The Glencoe left an hour or two later for North Sydney.”

In contrast, study of the issues from 1931 and 1932 produced no references to

storm activity or related delays. One article did comment upon heavy winds and snow at

Port aux Basques, but specifically mentioned that this did not affect the train schedules.

Study of the 1951-1952 issues revealed discussion of three storms sufficient to

cause damage and delay to railway traffic. The 29 June 1951 issue (p. 5) also reported

that,

“Two Men Badly Injured When Gale Blows Truck Over On The Highway.

St. Andrews – (Special) – An accident occurred on the Trans-Canada

Highway last Sunday when a truck turned over and two men were badly

injured. It is reported that the truck blew over during a severe southeast gale

which raged over the Valley that day. This is the first accident to occur on

the new highway.”

The low totals for delays recorded in the Western Star for these sample years

(between 0 and 4.5 per year for the three time periods) contrast with the annual average

of 7.1 delays per year noted by Sutherland et al. (1963). Improvements in engine design,

weather forecasting, and track maintenance would be expected to result in decreasing

total delays over time, under a regime of consistent climate. However, the low totals

more likely reflect a lack of reporting of incidents in the Western Star, rather than an

actual increase in the number of delays. Comparison of the 3 delays reported over the

1951-1952 period in the Western Star with the 7.1 yearly average noted between 1956

61

and 1962 suggests that under-reporting is more likely than a rapid increase in gale

frequency.

4.3 Canadian National Archives

Delays to rail traffic should primarily be recorded by the operators.

Consequently, inquiries were made to locate the CN archival material relevant to the

Newfoundland Railway, although previous efforts in this regard in conjunction with other

studies (e.g. Catto and Hickman, 2004; Catto, 2006b) had proven unsuccessful.

The CN archives are part of the Library and Archives Canada collection in

Ottawa (Record Group 30). The available records appear to consist of the Newfoundland

Railway Records for 1881, 1884, and 1904-1946, specifically including:

1) income and expenditure graph 1904-1946

2) tariffs and pay book 1915-1949

3) photographs and time table 1928

Initial investigation, facilitated by Library and Archives Canada staff, revealed no

material pertinent to delays of traffic, or to repairs of the line. The available records

concerned primarily tariffs, timetables, price books of supplies, and bills paid. Further

research would entail diligent analysis of all the materials, but the preliminary

investigations suggest that no archival materials pertaining to delays in traffic have been

maintained.

4.4 Other Archival Research

On advice from Library and Archives Canada, the Merrilees Transportation

Collection website, part of the Library and Archives Canada website

(http://www.collectionscanada.ca/trains/h30-6000-e.htm) was consulted, but yielded no

useful information. Further research in the archives of the Canadian Railroad Historical

Association, assisted by Josee Vallarand, archivist; the Provincial Archives of

Newfoundland and Labrador (including the Reid Collection); the Railway Coastal

Museum (of NL), assisted by Pam Correstine, former archivist; the AC Hunter Library;

and the City of St. John’s archives also proved unsuccessful.

62

Search at the Provincial Archives of Newfoundland and Labrador revealed a

single item of potential interest. File 321 “Reid Nfld Railway, Statements, 1902-1919”, in

addition to containing general statements of expenses such as maintenance of the entire

railway line, also contains a chart entitled “Data – Reid Newfoundland Railway” which

lists the miles of track, tons of freight carried, passengers carried, tie renewals, and other

data for 1902 to 1919. It also lists train derailments, for the entire line, between 1904 and

1919. Derailment totals varied from 13 (1908) to 94 (1918). Neither the chart nor

anything else in the file specifies the cause of the train derailments. The high totals,

however, do suggest that train derailments were common occurrences, and thus perhaps

were not comprehensively reported in newspapers at the time.

4.5 Summary

• At present, it is not possible to assess if any change has occurred in the

frequency of Wreckhouse Events, as judged from resultant delays to transportation,

over the historical record. The available archival data appears to be inadequate for

such an assessment.

• Recent events, documented by meteorological data, are discussed in Chapter 5.

Historically, Wreckhouse Events have been known and respected since the initial

railway construction period.

• The most recent cost estimates of disruption to highway traffic, summarized by

ADI Limited (1994), suggest that accidents cost ca. $236,000 annually, and delays

cost a further $1.9 million (as measured in 2006 dollars). Thus, the cost of

Wreckhouse Events to the provincial economy is in excess of $ 2.1 million

annually.

• This figure only includes insured losses and delays, and does not include delays

or damage to smaller, non-commercial vehicles, or costs involved with response to

accidents or road maintenance and repair.

63

5. Climate and Meteorological Analyses

Dale Foote and Dermot Kearney

5. 1. Introduction

Southwestern Newfoundland and the Cabot Strait are vital to maintaining a link

between the Island of Newfoundland and the rest of Canada. It is important that regular

ferry service into the harbour at Port aux Basques and traffic movement via the Trans-

Canada Highway be maintained. Highway traffic is often disrupted in the Wreckhouse

area due to severe localized winds and ferry traffic due to high winds and rough seas in

the Cabot Strait.

An analysis of the region’s wind and wave climatology over the past 35 to 50 years

has been used to determine if there are any significant changes in wind velocity, duration

and frequency and wave height. Several data sources were exploited:

1. Environment Canada’s National Meteorological Data Archives.

2. The AES40 North Atlantic Wind and Wave Climatology.

3. Environment Canada’s 100 Year (1901 - 2000) Tropical Cyclone Climatology for

cities.

The southwest coast of Newfoundland and the adjacent waters of the Cabot Strait are

located near the primary storm tracks of North America. Storms track south or southeast

of the Cabot Strait or to the west or northwest through the Gulf of St. Lawrence toward

Labrador or the Labrador Sea. Frequent and intense storm activity is often associated

with low pressure systems approaching the Cabot Strait from the south. This is primarily

due to the proximity of the North American landmass to the warm waters of the Gulf

Stream. When cold air originating over the continental landmass, clashes with warmer

moist air over the Gulf Stream, frequent cyclogenesis occurs. This makes the Gulf Stream

off the eastern seaboard of the United States the most active area for cyclogenesis in the

North Atlantic. When these storms move northeastward toward Atlantic Canada they can

rapidly intensify in a favourable oceanic environment.

64

The interaction of these weather systems with the high terrain over southwestern

Newfoundland can cause complex wind patterns, prevailing wind at Port aux Basques to

be easterly, which is opposite to that of most places in North America, and severe

downslope wind storms observed at Wreckhouse and St. Andrew’s.

Downslope wind storms are observed on the lee side of major mountain ranges

such as the Canadian Rockies and the European Alps. These storms also occur on much

smaller geographical scales as observed at Wreckhouse in southwestern Newfoundland

and les Suetes winds at Cheticamp on Cape Breton Island (Desjardins, 1995). They occur

when stably stratified air is forced to rise over a topographic barrier causing oscillations

or disturbances in the mean flow. Depending on the height and shape of the terrain as

well as the vertical wind and temperature structure of the lower atmosphere, different

wind regimes can result including the Wreckhouse and les Suetes winds. During these

events air is forced over a topographical barrier and upon reaching a supercritical flow

(Froude number >1), will accelerate on the lee side of the barrier. As the disturbance

descends and accelerates intense winds can be concentrated at the ground, analogous to a

hydraulic jump, after which the flow begins to adjust back to a subcritical flow (Froude

number <1). Results of numerical model simulations on the les Suetes in Cape Breton

indicate a hydraulic analogy is applicable to describing that wind phenomena.

Simulations of the Wreckhouse wind events would likely give a similar result.

In addition to the baroclinic synoptic scale storms that develop in the mid

latitudes and move through the study area, the Atlantic tropical weather season between

June and November, and the associated tropical storms and hurricanes, may affect the

Cabot Strait region. Most of these storms weaken as they move northward and enter

cooler waters and some transition into an extra-tropical low or post-tropical system.

Whether storms originate in the mid latitudes or tropical areas of the Atlantic

Ocean, the Cabot Strait region is prone to frequent storms with high winds, heavy

precipitation, large waves and storm surges.

65

5.2. Data Sources

5.2.1 Environment Canada’s National Meteorological Data Archive

Environment Canada (EC) maintains the archive of meteorological data in Canada

and this study uses data from several sites contained in the archive: Port aux Basques

(WZB), Burgeo (WBF), Stephenville (YJT), Sydney (YQY), Wreckhouse (XWR) and St.

Andrew’s. Observation site history discussions are given below.

Figure 5.1. Map showing the locations of wind measurement stations used in the

Southwest Newfoundland Cabot Strait Transportation Study.

The weather observations taken at Port aux Basques have a long and dynamic

history. The earliest reports date back to 1929 when the observations were taken at the

old Telegraph and Cable office. They were taken twice daily from an anemometer on top

of a 150 foot cliff on a 20 foot tower. There were some meteorological inspector reports

from the early 1930s but it appears the location of the observations remained at the

telegraph office. No inspector’s reports were found for most of the 1930s and all of the

1940s. In 1955 the station was moved to the Loran transmitting station approximately 1.5

66

miles southwest of the Canadian National Railway station in Port aux Basques. From

1955 until June 1966 there were no wind measurements taken in Port aux Basques. In

June of 1966 the U2A anemometer at St. Andrew’s was relocated to Port aux Basques on

a hill 56 feet ASL and placed on a 43 ft high tower. The hill sloped away to the ocean in

all directions except north and northeast where the terrain gradually rose to 200 feet ASL

approximately three quarters of a mile away. There were no changes to the observation

location until August 1983 when it was moved, 800 ft to the north northeast, to its current

location. On this site the anemometer was 32 metres higher on the hill and placed on a 10

meter tower.

Meteorological observations at the Stephenville Airport began in January 1953

when the American military operated the facility as Harmon Field. From 1953 to June

1966 there were no inspector’s reports available. The Americans operated two wind

monitoring systems located at each end of the 10,000 foot runway 09-27. In 1966 the

Department of Transport managed the airport and a new anemometer was temporary

installed on a 6 metre tower until later in the year, likely September (MSC

Meteorological Inspector Brian Hulan, personal communication). The anemometer

remains at the same location today.

Meteorological observations in Burgeo began in the summer of 1966 and continue

today at the original location. The height of the anemometer has remained constant at 10

metres however in the summer of 1969 the type was changed from a 45B to a U2A. The

anemometer rests on a 10 metre tower on small hill 75 ft above MSL which is 40 feet

above station elevation. The site is well exposed in all directions.

Sydney Airport, as with Port aux Basques and Stephenville, has an inconsistent

anemometer height and location history (see Karn, 2004). From June 1950 to August

1968 the anemometer location and height were changed several times. On August 31,

1968 the wind observations began at the present location and height.

Wreckhouse, located northwest of Port aux Basques in the Codroy Valley, is an

open and barren stretch of land where trains were once blown off their tracks by severe

local windstorms. Now it is usually large transport trucks that fall victim to these

powerful winds (see Chapter 4). The place name Wreckhouse has thus become

synonymous with the Wreckhouse wind phenomena.

67

The meteorological observations, taken at Wreckhouse, have been consistent in

location and height since January 2000. Between 1953 and 1966 the observations were

taken at St. Andrew’s and in the 1980s and early 1990’s at the nearby Starlight Lounge.

These sites are in different areas of the Codroy Valley and also experience the

Wreckhouse Effect.

5.2.2 AES40 North Atlantic Wind and Wave Climatology

The AES40 data set was developed at Oceanweather a specialized consulting firm

serving the coastal and ocean engineering communities with support from the Climate

Research Branch of Environment Canada. The hindcast involved the kinematic reanalysis

of all significant tropical and extra-tropical storms in the North Atlantic for the period

1954-2003. This data set has the advantage of being consistent in its temporal and spatial

resolution and with the height of its predicted wind. There was no need for

homogenization of the data and changes in the wind climatology will be attributable to

the forecast climatology.

5.2.3 Environment Canada 100 Year (1901 - 2000) Tropical Cyclone Climatology for

Cities

Environment Canada has analyzed 100 years of tropical cyclone tracks through

Atlantic Canada and recorded when these storms pass within 260 kilometres of 33

different locations. It was determined in 1986 by Barks and Richards of EC Atlantic that

tropical cyclones coming within 260 km of a location have a high probability of

producing gale force winds at that location.

We have decided to use the criteria of a tropical cyclone passing within 260

kilometres of Sydney or Port aux Basques to indicate when a tropical cyclone had the

greatest probability of impacting travel through the Cabot Strait or southwestern

Newfoundland.

68

Figure 5.2 Winter Storm Tracks

(A) Great Lakes Lows (B) Cape Hatteras/Cape Cod Lows

(C) Gulf of Mexico Lows

Joint Frequency Distribution

Burgeo, NL Complete History (1966-2004)

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

6.91

3.56 2.91

4.59

6.02

14.05

6.54

2.98

2.43 3.06

2.99

4.70

5.53

13.33

9.84

6.07

4.50

Wind Speed ( Kilometers Per Hour)

1 11 21 31 41 51

Figure 5.3 Wind speeds and directions, Burgeo

69

Joint Frequency Distribution

Port aux Basques, NL (1966-2004) Complete History

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

3.32

3.86 2.27

2.35

5.95

21.71

4.93

2.40

2.18 3.18

2.40

2.01

3.53

13.66

13.20

8.91

4.14

Wind Speed ( kmh)

1 11 21 31 41 51

Figure 5.4 Wind speeds and directions, Port aux Basques

Joint Frequency Distribution

Stephenville A, NL (1967-2004) Complete History

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

7.54

3.63 1.77

7.63

10.42

8.48

2.04

1.49

1.91 4.66

3.10

5.29 12.07

14.24

6.80

5.24

3.71

Wind Speed ( kmh)

1 11 21 31 41 51

Figure 5.5 Wind speeds and directions, Stephenville

70

Joint Frequency Distribution

Sydney A, NS (1967-2004) Complete History

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

3.36

7.99

3.13

2.96

2.44

3.43

2.44

2.77

3.56

9.31 10.44

11.49

9.28

12.20

6.06

4.61

4.54

Wind Speed ( kmh)

1 11 21 31 41 51

Figure 5.6 Wind Speeds and directions, Sydney

5.3 Analysis

5.3.1 Inhomogeneous data

Inhomogeneous data is the result of an inconsistent measurement history, i.e.

changes in location, measurement technique or instrumentation. As documented in the

site histories section some data used for this report has such inconsistencies. Two

methods of correcting for non homogenous data were considered for this report.

1. Using the ratio of monthly averages before and after a modification in the sampling

methodology to develop a correction factor to apply to one portion of the data. This

method works well when only one variable changes. When there are changes occurring

in climatological and sampling methods at the same time it is difficult to quantify what

percent of change is due to instrumentation.

71

2. Correction of the wind speed using the standard atmosphere approximation. This

method does not incorporate frictional effects, changes in horizontal location or

atmospheric profiles that deviate from the standard atmosphere, for example stable or

unstable lower level lapse rates.

A robust analysis of the magnitudes of the various components that may

contribute to a step change associated with an in-homogeneity is beyond the scope of this

study. Adjustment of monthly means based on the ratio of monthly averages before and

after a measurement change is susceptible to being confused with natural interannual

variability. Therefore any analysis on long-term trends will be confined to the portions of

the climatology with minimal change in observing routine and homogenization of the

wind data will not be attempted.

All the trend analysis values in the paper were obtained using the Kendall’s tau

test for trend and associated p-values.

5. 3.2 Historical measured winds

The prevailing wind over North America is generally westerly. However, local

topography has an influence on the wind direction at any given location. The individual

stations sites for our study area have a wide variety of mean annual wind speeds and

directions.

Burgeo and Port aux Basques are influenced by cold air dammed along the

mountains and funnelling from the fjords along the south coast of Newfoundland. These

topographical effects cause prevailing winds, especially during the summer, to be

easterly. Doherty (1991) described a favourable meteorological pattern that causes the

coastal easterlies often observed at Port aux Basques and Burgeo. “Flows from the south

or southwest combine with cold sea surface temperatures to produce very stable air in

low levels. The stable flow is blocked by the hills along the south coast of Newfoundland

resulting in easterly surface winds along the coast”. (Doherty, 1991). Burgeo has a high

occurrence of easterly winds especially during the warmer months. Due to higher

topography near Port aux Basques than Burgeo, Port aux Basques experiences more

72

episodes of easterly winds. These terrain effects will likely have an impact in determining

any long term trends in the wind climatology of the southwest coast.

Stephenville Airport also has strong topographical forcing on its prevailing winds

but not as obvious as Burgeo or Port aux Basques. St. Georges Bay lies directly to the

southwest of the observing station causing the prevailing southwesterly winds to blow

relatively unimpeded. However the mountains that lie to the north and east of the

observing station increase the frequency of easterly winds in two ways.

• Redirection of synoptic scale flow ahead of an approaching low pressure system

from the south or southeast direction to the east to northeast direction.

• Katabatic/drainage winds can develop and flow down the mountainsides under

clear skies and a weak synoptic scale isobar pattern.

Sydney Airport’s prevailing wind direction is from the southwest quadrant with

the wind blowing between south and west 54% of the time. The remaining 46% is spread

relatively evenly between west northwest and south southeast. Sydney Airport is the least

topographically influenced observing station included in this study (Karn, 2004).

5.3.2.1 Yearly Mean Wind Speeds

Station Trend (kph/Year) N p-value**

Burgeo (WBF) -0.14 35 <0.001

Port aux Basques (WZB)

+0.078 38 0.003

Stephenville (YJT)

+0.025 38 0.208

Sydney (YQY) -0.13 36 <0.001

** Significant at 5% level. Table 5.1 Yearly mean wind speeds

Yearly mean wind speeds were calculated for all for observing stations by

summing all hourly observations for a calendar year and dividing by the number of

observations. The results are presented in Table 5.1. All statistical significance tests

were conducted at the 5% level.

73

Figure 5.7 Yearly average wind speed, Burgeo

Wbf Yearly Average Wind Speed

0

5

10

15

20

25

30

35

40

1960 1970 1980 1990 2000 2010

Year

Kp

h

Figure 5.8 Yearly average wind speed, Port-aux-Basques

Wzb Yearly Average Wind Speed

0

5

10

15

20

25

30

35

40

1960 1970 1980 1990 2000 2010

Year

Kp

h

Figure 5.9 Yearly average wind speed, Stephenville

Yjt Yearly Average Wind Speed

0

5

10

15

20

25

30

35

40

1960 1970 1980 1990 2000 2010

Year

Kp

h

74

Figure 5.10 Yearly average wind speed, Sydney

Yqy Yearly Average Wind Speed

0

5

10

15

20

25

30

35

40

1960 1970 1980 1990 2000 2010

Year

Kp

h

The results for Burgeo and Sydney Airport indicate a statistically significant

decreasing trend while Port aux Basques has a statistically significant increasing trend.

The p-values for these trend lines were very small. The slight increasing trend at

Stephenville is not statistically significant.

5.3.2.2 Yearly Maximum Wind Speeds

Station Trend (kph/Year) N p-value**

Burgeo (WBF) <.001 35 0.81

Port aux Basques (WZB)

+0.50 38 0.023

Stephenville (YJT)

0 38 0.69

Sydney (YQY) -0.40 36 0.002

** Significant at 5% level. Table 5.2. Yearly maximum wind speeds

Yearly maximum wind speeds were determined for the stations by selecting the

maximum wind speed for a calendar year. The results are presented in Table 5.2. All

significance tests were conducted at the 5% significance level.

The overall trends for yearly maximum wind speeds in table 2 are consistent with

the yearly mean wind speeds in table 1 where the results for Burgeo and Sydney Airport

indicated a decreasing trend while Port aux Basques had an increasing trend. However,

75

Figure 5.11 Yearly maximum wind speed, Burgeo

Wbf Yearly Maximum Wind Speed

0

20

40

60

80

100

120

140

1960 1970 1980 1990 2000 2010

Year

Kp

h

Figure 5.12 Yearly maximum wind speed, Port aux Basques

Wzb Yearly Maximum Wind Speed

0

20

40

60

80

100

120

140

1960 1970 1980 1990 2000 2010

Year

Kp

h

Figure 5.13 Yearly maximum wind speed, Stephenville

Yjt Yearly Maximum Wind Speed

0

20

40

60

80

100

120

140

1960 1970 1980 1990 2000 2010

Year

Kp

h

76

Figure 5.14 Yearly maximum wind speed, Sydney

Yqy Yearly Maximum Wind Speed

0

20

40

60

80

100

120

140

1960 1970 1980 1990 2000 2010

Year

Kp

h

the slight decreasing trend at Burgeo is not considered to be statistically significant at the

5% level whereas Sydney’s decreasing trend is statistically significant with a p-value of

.002. There was no trend observed in Stephenville's data however it is not statistically

significant at the 5% level. Both Port aux Basques and Sydney’s trends are in the

opposite direction but both are statistically significant at the 5% level.

5.3.2.3 Annual Wind Speed Category Percentages

Annual wind speeds are divided into four categories: <19 knots, strong [19 to 33

knots], gale force [34 to 47 knots] and storm force [48 to 63]. Sustained hurricane force

winds rarely happen so they have not been included in these figures. However when

sustained storm force winds are measured they are often accompanied by hurricane force

wind gusts.

Table 5.3 contains results of the statistical analysis on the annual wind data using

the Kendall Tau test for trend. The results show that the majority of the change that has

been analyzed in the annual mean wind speeds is a result of a redistribution of winds in

the <19 knots and the strong categories. Changes in gale force winds have had less

obvious trends and they tend to be an order of magnitude smaller. Trends in storm force

winds are 2 orders of magnitude smaller.

77

Table 5.3.

Station/Category Trend (%/Year) N p-value**

Burgeo < 19 knots

+0.250 35 <0.001

Burgeo Strong

-0.233 35 <0.001

Burgeo Gale force

-0.011 35 0.120

Burgeo Storm force

0 35 0.629

Port aux Basques < 19 knots

-0.19 38 0.066

Port aux Basques Strong

+0.147 38 0.019

Port aux Basques Gale force

+0.043 38 0.007

Port aux Basques Storm force

+0.005 38 0.001

Stephenville < 19 knots

-0.021 38 0.249

Stephenville Strong

+0.025 38 0.247

Stephenville Gale force

< -0.001 38 0.939

Stephenville Storm force

< -0.001 38 0.628

Sydney < 19 knots +0.15 36 <0.001

Sydney Strong

-0.15 36 <0.001

Sydney Gale force

-0.005 36 0.001

Sydney Storm force

Insufficient Data

** Significant at 5% level. 5.3.2.4 Seasonal Wind Speed Category Percentages

Seasonal wind speed category graphs are presented in Appendix B and divided

into four categories: <19 knots, strong [19 to 33 knots], gale force [34 to 47 knots] and

storm force [48 to 63]. Sustained hurricane force winds have not been included. The

seasonal wind categories were calculated using climate data counts of the number of

78

times a wind speed fell within the desired category. The seasonal totals were obtained

and divided by the total number of observations and multiplied by 100 to obtain the

percentage of occurrence. Tables 5.4A to 5.4D contain results of the statistical analysis

on the seasonal wind data using the Kendall Tau test for trend. The majority of change is

due to the redistribution of winds between the <19 knots and the strong wind categories.

This is particularly evident in the summer and fall and is due to the seasonal cycle to

lower wind speeds. In the warmer months the higher wind speed categories have

insufficient data to determine a trend for most stations

5.3.3 Comments on the Seasonal Wind Speed Categories

The analysis for Burgeo indicates a statistically significant increasing trend in the

<19 knots category and a statistically significant decreasing trend for most of the higher

wind categories for all months. The exceptions occur in the spring and fall where the gale

wind category has a decreasing trend that is not statistically significant.

The analysis for Port aux Basques indicates a decreasing trend in the <19 knots

category and an increasing trend for the higher wind speed categories. The gale force

wind category indicates a statistically significant increasing trend for all seasons.

When the trend analysis is applied to the data between 1990 and 2002 for Port aux

Basques, there is a statistically significant decreasing trend for the <19 knots category

and a statistically significant increasing trend for the higher wind categories for most

seasons.

The analysis for Stephenville indicates a decreasing trend for the <19 knots

category and increasing trend for the higher wind speed categories however the trends in

all months are not statistically significant at the 5% level.

The analysis for Sydney indicates there is a statistically significant increasing

trend in the <19 knots category and a statistically significant decreasing trend for the

higher wind categories for all months.

79

Table 5.4A. Seasonal Trends DJF

Station/Category Trend

(%/Season)

N P-Value**

Burgeo < 19 knots

+.242 35 0.010

Burgeo Strong

-0.224 35 0.009

Burgeo Gale force

-0.038 35 0.10

Burgeo Storm force

0.00 35 0.579

Port aux Basques < 19 knots

-0.151 38 0.289

Port aux Basques Strong

+0.102 38 0.320

Port aux Basques Gale force

+0.10 38 0.024

Port aux Basques Storm force

+0.002 38

0.260

Stephenville < 19 knots

-0.082 38 0.204

Stephenville Strong

+0.079 38 0.148

Stephenville Gale force

Insufficient data

Stephenville Storm force

Insufficient data

Sydney < 19 knots

+0.224 36 0.001

Sydney Strong

-0.212 36 0.002

Sydney Gale force

-0.012 36

<0.001

Sydney Storm force

Insufficient data

** Significant at 5% level.

80

Table 5.4B. Seasonal Trends MAM

Station/Category Trend

(%/Season)

N P-Value**

Burgeo < 19 knots

+0.254 35 0.002

Burgeo Strong

-0.256 35 0.001

Burgeo Gale force

-0.008 35 0.56

Burgeo Storm force

Insufficient data

Port aux Basques < 19 knots

-0.205 38 0.105

Port aux Basques Strong

+0.152 38 0.133

Port aux Basques Gale force

+0.027 38 0.036

Port aux Basques Storm force

Insufficient data

Stephenville < 19 knots

-0.012 38 0.831

Stephenville Strong

+0.008 38 0.870

Stephenville Gale force

+0.007 38 0.870

Stephenville Storm force

Insufficient data

Sydney < 19 knots

+0.196 36 <0.001

Sydney Strong

-0.196 36 <0.001

Sydney Gale force

Insufficient data

Sydney Storm force

Insufficient data

** Significant at 5% level.

81

Table 5.4C. Seasonal Trends JJA

Station/Category Trend

(%/Season)

N P-Value**

Burgeo < 19 knots

+0.129 35 <0.001

Burgeo Strong

-0.107 35 0.001

Burgeo Gale force

Insufficient data

Burgeo Storm force

Insufficient data

Port aux Basques < 19 knots

-0.010 38 0.025

Port aux Basques Strong

+0.100 38 0.021

Port aux Basques Gale force

Insufficient data

Port aux Basques Storm force

Insufficient data

Stephenville < 19 knots

-0.003 38 0.744

Stephenville Strong

+0.002 38 0.890

Stephenville Gale force

Insufficient data

Stephenville Storm force

Insufficient data

Sydney < 19 knots

+0.066 36 <0.001

Sydney Strong

-0.064 36 <0.001

Sydney Gale force

Insufficient data

Sydney Storm force

Insufficient data

** Significant at 5% level.

82

Table 5.4D. Seasonal Trends SON

Station/Category Trend

(%/Season)

N P-Value**

Burgeo < 19 knots

+0.274 35 <0.001

Burgeo Strong

-0.244 35 <0.001

Burgeo Gale force

-0.012 35 0.382

Burgeo Storm force

Insufficient data

Port aux Basques < 19 knots

-0.156 38 0.084

Port aux Basques Strong

+0.117 38 0.132

Port aux Basques Gale force

+0.032 38 0.018

Port aux Basques Storm force

Insufficient data

Stephenville < 19 knots

-0.009 38 0.660

Stephenville Strong

+0.012 38 0.610

Stephenville Gale force

Insufficient data

Stephenville Storm force

Insufficient data

Sydney < 19 knots

+0.149 36 0.002

Sydney Strong

-0.145 36 0.002

Sydney Gale force

Insufficient data

Sydney Storm force

Insufficient data

** Significant at 5% level.

83

Figure 5.15

Burgeo Annual Wind Speed Categories

% Occurrence vs. Year

0.01

0.1

1

10

1001970

1973

1976

1979

1982

1985

1988

1991

1994

1997

2000

2003

Year

% < 19

% Strong

% Gale

% Storm

Figure 5.16

Port aux Basques Annual Wind Speed Categories

% Occurrence vs. Year

0.01

0.1

1

10

100

1967

1970

1973

1976

1979

1982

1985

1988

1991

1994

1997

2000

2003

Year

% < 19

% Strong

% Gale

% Storm

84

Figure 5.17

Stephenville Airport Annual Wind Speed Categories

% Occurrence vs. Year

0.01

0.1

1

10

100

1967

1970

1973

1976

1979

1982

1985

1988

1991

1994

1997

2000

2003

Year

% < 19

% Strong

% Gale

% Storm

Figure 5.18

Sydney Airport Annual Wind Speed Categories

% Occurrence vs. Year

0.01

0.1

1

10

100

1969

1972

1975

1978

1981

1984

1987

1990

1993

1996

1999

2002

Year

% < 19

% Strong

% Gale

% Storm

85

5.4 Transportation Interruption Index

Quantitative analysis of long-term trends in mean, maximum and categorical

winds does not address the issue of wind event duration. To investigate the possibility of

long-term trends in wind event duration, and the relative strength of those events, we

have developed a Transportation Interruption Index to help quantify the potential a given

year had to interrupt transportation through the Cabot Strait. In our index a wind event is

defined as lasting three hours and having a minimum wind speed of 63 kph. The

maximum wind speed is then multiplied by the mean wind speed and the resulting

product is multiplied by the duration. The total is then divided by a scaling factor of

11,907, the minimum value attainable from the above calculation. Therefore the

minimum value of the event, TIIe, is 1. The TIIe for the calendar year are then summed to

create the TII. This formula is also described below:

TIIe = ((Max Wind Speed)*(Mean Wind Speed)*(Duration))/11907

TII = ∑TIIe

The TII for Port aux Basques, Burgeo, Stephenville and Sydney are shown in

Figures 5.19 to 5.22. The results for Port aux Basques are further broken down to their

westerly and easterly components in Figures 5.23 and 5.24). The trend analysis for the

TII is shown in Table 5.5.

It is interesting to note the high interannual variability in the TII westerly

component compared to the relatively benign easterly. This graph shows quantitatively

that although the easterly winds are of concern for the Wreckhouse area, it is the

westerlies that have the greatest potential to produce stormy conditions in the Cabot

Strait.

86

Yearly Port aux Basques TII

(1966, 1997 & 1998 removed)

0

50

100

150

200

250

1960 1970 1980 1990 2000 2010

Figure 5.19

Yearly Burgeo TII

(1967-69 removed)

0

50

100

150

200

250

1960 1970 1980 1990 2000 2010

Figure 5.20

Yearly Stephenville TII

0

50

100

150

200

250

1960 1970 1980 1990 2000 2010

Figure 5.21

Yearly Sydney TII(1966-68 removed)

0

50

100

150

200

250

1960 1970 1980 1990 2000 2010

Figure 5.22

87

Figure 5.23 Transportation Interruption Index for Wreckhouse, Westerly winds

Yearly WZB TI Index - Westerly Wind

(1966, 1997 &1998 removed)

0

50

100

150

200

250

1960 1970 1980 1990 2000 2010

Year

TII

Figure 5.24 Transportation Interruption Index for Wreckhouse, Easterly Winds

Yearly WZB TI Index - Easterly Wind

(1966, 1997 &1998 removed)

0

50

100

150

200

250

1960 1970 1980 1990 2000 2010

Year

TII

88

Table 5.5.

Station Trend (per Year) N p-value**

Burgeo -1.13 35 0.002

Burgeo++

-0.011 33 0.988

Port aux Basques +2.40 36 0.004

Stephenville 0 38 0.850

Sydney -0.139 36 < 0.001

** Significant at 5% level. Burgeo++ - For comparison the extreme data for 1971 & 72 has been removed.

5.5 Wreckhouse Area

Port aux Basques and Wreckhouse are geographically close. However, the

topography of southwestern Newfoundland impacts the winds measured at both stations

differently, for identical synoptic weather patterns. Chapter 4 contains information on

past wind studies of the area that highlight the importance of the Wreckhouse wind.

For years the residents of southwestern Newfoundland and forecasters at the former

Newfoundland Weather Centre have known that a relationship exists between easterly

winds measured at Port aux Basques and the downslope winds present in the Wreckhouse

area. This relationship is shown in a graphical manner in Figures 5.25 and 5.27. The data

plotted are hourly observations and the corresponding wind speeds at both observing sites

from 2000 to 2004. The wind speeds would be very close in magnitude without the

topographical effect caused by the mountains and coastline and the data would be plotted

with a slope near one.

Figure 5.25 shows the relationship between the winds at Port aux Basques and

Wreckhouse for wind directions northeast to southerly when the mountains along the

south coast induce a downslope wind at Wreckhouse. This leads to the general

relationship that the winds at Wreckhouse are stronger than those at Port aux Basques for

these directions.

Figure 5.26 is for the westerly direction and shows that the winds at Port aux Basques

generally blow stronger than the winds at Wreckhouse for a similar synoptic scale

weather pattern. This is likely caused by the funnelling and channelling of westerly

89

Figure 5.25 Relationship between Port aux Basques (WZB) and Wreckhouse (XWR) winds, easterly direction

WZB vs XWR Wind Speed

Direction 060-180

0

20

40

60

80

100

120

140

0 20 40 60 80 100 120 140

Wzb Windspeed (kph)

Xw

r W

ind

sp

eed

(kp

h)

Figure 5.26 Relationship between Port aux Basques (WZB) and Wreckhouse (XWR) winds, westerly direction

WZB vs XWR Wind Speed

Direction 230-340

0

20

40

60

80

100

120

140

0 20 40 60 80 100 120 140

Xw

r W

indspeed (kph)

Wzb Windspeed (kph)

90

Figure 5.27 Relationship between Port aux Basques and Wreckhouse winds, easterly storm peaks

Figure 5.28 Inferred Wreckhouse Peak Winds, 1966-2004

C4

xwr wind

2002-04

1999-01

1996-98

1993-95

1990-92

1987-89

1984-86

1981-83

1978-80

1975-77

1972-74

1969-71

1966-68

140

130

120

110

100

90

80

Inferred Wreckhouse Peak Wind

Wzb vs Xwr Easterly Storm Peaks

with Wzb Speed >=57 kph

y = 0.8139x + 42.165R

2 = 0.6123

0

20

40

60

80

100

120

140

0 20 40 60 80 100 120 140

Wzb Wind Spd (kph)

Xw

r W

ind

Sp

d (

kp

h)

91

winds through the Cabot Strait and the exposure of the Port aux Basques area to this wind

direction.

Figure 5.27 displays the relationship differently. Wreckhouse events were identified

and the maximum hourly winds at the Port aux Basques observing station was paired

with the maximum hourly winds at the Wreckhouse observing station. It should be noted

that this is a relationship between hourly winds. There were no gusts analyzed, no special

observations included and some outliers were removed.

The linear regression line equation was used to infer the history of Wreckhouse winds

from the Port aux Basques wind observation data and the results are presented in Figures

5.28 and 5.29. They show considerable inter-annual and inter-decadal variability in the

frequency of Wreckhouse events and the maximum sustained wind expected. However

there are no long term increasing or decreasing trends present in the data.

Wreckhouse Event Total Yearly Hours (1966, 1997 &1998 Removed)

0

100

200

300

400

500

1960 1970 1980 1990 2000 2010

Year

Ho

urs

Figure 5.29 Wreckhouse Wind event total yearly hours. See Chapter 7 for further

discussion.

92

5.6 Wreckhouse Area Wind Climatology

St. Andrew’s 1953-May 1966

Wreckhouse 2000-2004

In Figures 5.30 to 5.35 a Wreckhouse wind event is defined as having occurred if

the mean wind speed reached 75 km/h or more and from directions 080o clockwise to

170o inclusive. Multiple Wreckhouse events could occur for the same synoptic system if

the duration of an event was interrupted by 3 or more hours of lighter wind speeds.

Wreckhouse events have been observed to last for 16 hours at St. Andrew’s and 22 hours

at Wreckhouse. The figures show most Wreckhouse events last longer than 3 hours. Gust

data has not been included.

The annual distribution of Wreckhouse events at St. Andrews’ shows much

variability ranging from 4 events in 1955 and 1957 to 15 in 1963. Similar variability is

observed in six years of data from the Wreckhouse observing station. There were also

more events observed at Wreckhouse however these stations should not be compared.

The St. Andrew’s data was collected from manned observations and the Wreckhouse data

from an automatic weather station. Both stations are separated by several kilometres and

used different wind measuring instruments and recording periods.

The monthly distribution of Wreckhouse events is very similar at both locations

and this is expected. Wreckhouse events occur most often in the winter and early spring

with little or no events during the summer. This is due to the strengthening of the jet

stream and synoptic weather systems in the winter and a weakening during the summer.

Meteorologists in Atlantic Canada have known that Wreckhouse windstorms

mainly occur with southeast winds ahead of low pressure systems that track west or

northwest of the Cabot Strait and much less frequently for ones that track to the southeast

(Figure 5.36). This is also evident from the Wreckhouse wind study (Sutherland, 1963)

that found 20 percent of the Wreckhouse windstorms occurred when low pressure

systems tracked southeast of the Cabot Strait and 80 percent when they travelled to the

west or northwest.

93

Figure 5.30

Yearly Distribution of Wreckhouse Events at St. Andrew's

Maximum Mean Wind Speed >= 75 km/h

0

2

4

6

8

10

12

14

161

95

3

19

54

19

55

19

56

19

57

19

58

19

59

19

60

19

61

19

62

19

63

19

64

19

65

19

66

Year

#E

ve

nts Duration <3 hrs

Duration >=3 hrs

Figure 5.31

Monthly Distribution of Wreckhouse Events 1953-1966 at

St. Andrew's Maximum Mean Wind Speed >=75 km/h

0

5

10

15

20

25

Ja

n

Fe

b

Ma

r

Ap

r

Ma

y

Jun

e

Ju

ly

Au

g

Se

pt

Oct

Nov

De

c

Month

# E

ve

nts

94

Figure 5.32

Total # Observations at St. Andrew's

1953-1966

0

2000

4000

6000

8000

10000

195

3

195

4

195

5

195

6

195

7

195

8

195

9

196

0

196

1

196

2

196

3

196

4

196

5

196

6

Year

Figure 5.33

0

5

10

15

20

25

30

35

2000 2001 2002 2003 2004

Year

#E

ve

nts

< 3 hrs

>= 3 hrs

Yearly Distribution of Wreckhouse Events at

Wreckhouse 2000-2004

Maximum Mean Wind Speed >= 75 km/h

95

Figure 5.34

Monthly Distribution of Wreckhouse Events at

Wreckhouse 2000-2004

0

5

10

15

20

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Month

# E

ven

ts

Figure 5.35

Total # Observations at Wreckhouse 2000-2004

7200

7400

7600

7800

8000

8200

8400

8600

8800

9000

2000 2001 2002 2003 2004

Year

96

Figure 5.36 Wreckhouse Wind Event Storm Tracks

Tables 5.6A and 5.6B highlight the significance of the Wreckhouse winds in the

Codroy Valley. All maximum annual average wind speeds (all directions), at St.

Andrew’s and Wreckhouse, occurred during Wreckhouse wind events.

Table 5.6A.

Maximum Annual Average Wind Speed

St. Andrew’s, NL 1953-1966

YEAR

Wind Speed (km/h)

(1 Hour Average)

Direction

1953 140o@ 97

1954 140o@ 97

1955 140o@ 100

1956 160o@ 105

1957 110o@ 129

1958 160o@ 116

1959 140o@ 145

1960 180o@ 121

1961 110o@ 145

1962 140o@ 97

1963 140o@ 129

1964 140o@ 126

1965 110o@ 113

1966 Jan-May 110o@ 89

97

Table 5.6B.

Maximum Annual Average Wind Speed

Wreckhouse, NL 2000-2005

YEAR

Wind Speed (km/h)

(1 Hour Average)

Direction

2000 100o@ 128

2001 110o@ 120

2002 120o@ 120

2003 130o@ 126

2004 090o@ 122

2005 100o@ 110

5.7 AES40 Model data

Figures 5.37 and 5.38 show the annual average and maximum wind speeds,

respectively, for the AES40 data point. The results of the statistical analysis are presented

in table 5.7. There is a statistically increasing trend for the average yearly wind over the

period 1955 to 2002. When the 1967 to 2002 time period, roughly the same time that the

data from the observation network is being considered, the statistical significance is lost

and the trend resembles the one in the Stephenville Airport dataset.

Table 5.7.

AES40 Trend

(kph/Year)

N p-value **

Yearly Average (1955-2002) +0.028 48 0.032

Yearly Maximum (1955-2002) +0.106 48 0.080

Yearly Average (1967-2002) +0.007 36 0.733

Yearly Maximum (1967-2002) -0.004 36 0.859

** Significant at 5% level. Figure 5.39 shows the mean daily maximum wind for winter (DJF) and (figure 10.6)

shows the annual frequency of gale force or greater winds. The analysis is given in Table

5.7A. Figures 5.40 and 5.41 show box plots of winter daily maximum winds and monthly

maximum winds, respectively. The lack of statistical significance shown in Table 5.7A

98

Figure 5.37

AES 40 Yearly Average Wind Speed

0

5

10

15

20

25

30

35

40

1950 1960 1970 1980 1990 2000 2010

Year

Kp

h

Figure 5.38

AES 40 Yearly Maximum Wind Speed

0

20

40

60

80

100

120

140

1950 1960 1970 1980 1990 2000 2010

Year

Kp

h

99

Figure 5.39

AES 40 Mean Winter Daily Max Wind

30

35

40

45

50

55

1950 1960 1970 1980 1990 2000 2010

Year

kp

h

Figure 5.40

Year

Daily Max (kph)

2003

2002

2001

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

1988

1987

1986

1985

1984

1983

1982

1981

1980

1979

1978

1977

1976

1975

1974

1973

1972

1971

1970

1969

1968

1967

1966

1965

1964

1963

1962

1961

1960

1959

1958

1957

1956

1955

100

80

60

40

20

0

AES40 Winter Daily Max Winds

100

Figure 5.41

Year

Monthly Max W

ind Speed

2002/03

2000/01

1998/99

1996/97

1994/95

1992/93

1990/91

1988/89

1986/87

1984/85

1982/83

1980/81

1978/79

1976/77

1974/75

1972/73

1970/71

1968/69

1966/67

1964/65

1962/63

1960/61

1958/59

1956/57

1954/55

100

90

80

70

60

50

40

30

Monthly Max Winds

Figure 5.42

AES 40 Count of > 34 Kts(1954 & 2003 removed)

0

10

20

30

40

50

1950 1960 1970 1980 1990 2000 2010

Year

Co

un

t

can also be seen in these plots with considerable interannual variability but no clear trend

is evident.

101

Table 5.7a.

AES40 Wind

Trends

Trend

(+/-)

N p-value **

Mean Winter (DJF) Maximum Wind

+ 48 0.294

Annual Frequency of Wind >= 34 knots

+ 48 0.227

AES40 yearly wind speeds were sorted into marine wind categories for a calendar year.

The results are presented in Figure 5.43 and the statistical analysis is presented in table

5.8. All significance tests were conducted at the 5% significance level.

Figure 5.43

AES40 Annual Wind Speed Categories

0.1

1

10

100

1955

1957

1959

1961

1963

1965

1967

1969

1971

1973

1975

1977

1979

1981

1983

1985

1987

1989

1991

1993

1995

1997

1999

2001

Year

Perc

en

tag

e

%<strong

% strong

%gale

Table 5.8.

Wind Category

1955-2002

Trend

(kph/Year)

N p-value **

< Strong (19 knots) -0.080 48 0.032

Strong (19 to 33 knots) +0.080 48 0.029

>= Gale (34 knots) +0.006 48 0.380

** Significant at 5% level.

102

The result for the strong category indicates a statistically significant increasing

trend while the less than strong category has a statistically significant decreasing trend.

The p-values for these regression lines were 0.029 and 0.032 respectively. The increasing

trend for the gale force or greater category is not statistically significant at the 5% level.

This analysis indicates an overall decrease in weaker winds but it does not translate into

an increased frequency of more extreme winds.

An analysis of the AES40 wave data set for the Cabot Strait is presented in

Figures 5.44 and 5.45. The results are summarized in Table 5.9.

Figure 5.44

Yearly Peak Sig and Max Wave Height

(1954 & 2003 removed)

0

5

10

15

20

1950 1960 1970 1980 1990 2000 2010

Year

He

igh

t (m

)

Sig Wave

Max Wave

Figure 5.45

Yearly Count of Wave >= 4 m(1954 $ 2003 removed)

0

10

20

30

40

50

60

70

1950 1960 1970 1980 1990 2000 2010

Year

Co

un

t

103

Table 5.9.

Yearly AES40 wave

value 1955-2002

TREND

cm/yr

N p-value **

Peak Maximum wave height

+0.070 48 0.002

Peak Significant wave height

+0.035 48 0.005

Count of Waves >= 4 metres

0.00 48 0.850

** Significant at 5% level. The result for the peak yearly maximum and significant wave height categories

indicate a statistically significant increasing trend. The annual count of significant wave

heights greater than 4 metres show no trend that is not statistically significant.

5.8 Environment Canada 100 Year (1901 - 2000) Tropical Cyclone Climatology

Few meteorological phenomena capture the attention of meteorologists, the media

and the general public as hurricanes do. Images of these storms charting their way

through the Gulf of Mexico causing the evacuation of millions of people have a way of

leaving an impression. Eventually many of these storms make their way to Atlantic

Canada near south-western Newfoundland (Figure 5.46) where they can take many

forms. The most common scenarios for a tropical cyclone that passes through Atlantic

Canada is to weaken as it crosses eastern North America or weaken over the cold waters

of the North Atlantic. In these cases the winds will diminish considerably below

hurricane strength and the widespread torrential rain will diminish to areas of heavy

showers. Hurricane Ophelia (2005) is a good example.

However there are scenarios when these storms maintain their strength and make

landfall as full-fledged hurricanes or interact with processes in the upper atmosphere that

will cause intensification. These storms are accompanied by torrential rains, hurricane

force winds and damaging waves and storm surge. In October 2000 hurricane Michael

made landfall on the south coast of Newfoundland as a category one hurricane.

104

Figure 5.46 Tropical cyclone tracks

Environment Canada has analyzed 100 years of tropical cyclone tracks as they pass

through Atlantic Canada and recorded when these storms pass within 260 kilometres of

33 different locations. We have decided to use the criteria of a tropical cyclone passing

within 260 kilometres of Sydney or Port aux Basques to determine when a tropical

cyclone might impact travel through the Cabot Strait or southwestern Newfoundland. The

results are shown in Figures 5.47 and 5.48. Figure 5.47 shows the monthly frequency

distribution of tropical cyclone activity with September being the most active month.

Figure 5.48 shows the annual frequency of tropical cyclone activity over 100 years. An

active year consists of 3 storms and many years in which no storms pass within 260 km

of the Cabot Strait area. There is no obvious trend in tropical cyclone activity over the

study area.

105

Figure 5.47

Figure 5.48

Tropical Cyclone Frequency by Year

Port-aux Basques and Sydney

260 km radius

1901-2000

0.00

1.00

2.00

3.00

4.00

19

01

19

03

19

05

19

07

19

09

19

11

19

13

19

15

19

17

19

19

19

21

19

23

19

25

19

27

19

29

19

31

19

33

19

35

19

37

19

39

19

41

19

43

19

45

19

47

19

49

19

51

19

53

19

55

19

57

19

59

19

61

19

63

19

65

19

67

19

69

19

71

19

73

19

75

19

77

19

79

19

81

19

83

19

85

19

87

19

89

19

91

19

93

19

95

19

97

19

99

5.9 Discussion: Inhomogeneous Data

In order to draw as much trend information from the data as possible the

inhomogeneities in the data, documented and perhaps otherwise, need to be considered.

To determine if a change in wind speed is related to changes in the physical environment

or changes in the observation history it is helpful to determine when the wind speed

changed. If the change corresponds with an observation method or location change it

raises the possibility that the increase or decrease in wind speed is not entirely natural but

Tropical Cyclone Frequency by Month

Port-aux-Basques and Sydney

260 km radius

1901-2000

0

10

20

30

Jan Feb Mar Apr May June July Aug Sep Oct Nov Dec

106

caused by the non-homogeneous nature of the measurements. A double mass curve is an

arithmetic plot of the accumulated values of observations of two variables that are paired

in time and thought to be related. As long as the relationship remains constant the double-

mass curve will appear as a straight line; a deviation denotes the timing of a change1. The

annual percentages of strong winds for our observation sites where inhomogeneities exist

are compared to the annual percentages of strong winds AES40 data set where

inhomogeneities do not exist. Results are displayed in Figures 5.49 to 5.53.

Figure 5.49 shows the accumulated annual percentages of strong winds for all 4

sites plus the AES40 site. Note the relative consistent slope of the AES40 data when

compared to the historical climatological data.

Figure 5.50, Burgeo indicates changes in 1975, 1979, 1985 and 1993. However

none of these inflexion points correspond to dates of change in the site history.

Figure 5.51, Port aux Basques indicates changes in 1972, 1980, 1984, 1992 and

1997. There were changes to site history in 1983, 1991 and 1998.

Figure 5.52, Sydney indicates changes in 1971, 1978, 1982, 1989 and 1993 and

possible more.

Figure 5.53, Stephenville indicates slight changes in 1974, 1988, 1994 and 1996.

None of these inflexion points correspond to dates of changes as indicated by the

inspector reports.

From the analysis of the double mass curve it appears the wind data observed at

Stephenville has the highest probability of being the least affected by the documented

inhomogeneities in its observational history. The changes present in its curve do not

correspond to the documented changes in observing method or location. The observations

at Burgeo also have a high probability of not being affected by the documented changes

in observing method but there seems to be a significant and persistent reduction in the

percentage of strong winds beginning in the mid 1980s. This could be a natural reduction

in the local winds or the result of changes in the physical environment that are not

documented in the inspector’s reports. The changes in the Port aux Basques curve

correspond to changes in observing method. Changes in Sydney’s curve are so dramatic it

1 http://amsglossary.allenpress.com/glossary

107

Figure 5.49

Accumulated % Strong Winds VS. Year

0

100

200

300

400

500

600

700

800

900

1000

1960 1970 1980 1990 2000 2010Year

%

AES40

YJT

WBF

WZB

YQY

Figure 5.50

Double Mass Curve:% Strong Winds

WBF VS. AES 40

0

100

200

300

400

500

600

700

800

0 200 400 600 800 1000

108

Figure 5.51

Double Mass Curve:% Strong Winds

WZB VS. AES 40

0

100

200

300

400

500

600

700

0 200 400 600 800 1000

Figure 5.52

Double Mass Curve:% Strong Winds

YQY VS. AES 40

0

50

100

150

200

250

300

0 200 400 600 800 1000

Figure 5.53

Double Mass Curve:% Strong Winds

YJT VS. AES 40

0

50

100

150

200

250

300

0 200 400 600 800 1000

109

is hard to imagine such changes occurring naturally in such a short period of time.

Accordingly the Stephenville and Burgeo historical data will be given the greatest weight

when determining the presence of long term trends.

5.9 Summary of Overall Trends

Our analysis of the wind regime for southwestern Newfoundland and the Cabot

Strait included an analysis of historical and model derived climatological data, model

derived wave statistics and data presented in Environment Canada's 100 year tropical

cyclone climatology for the last 35 to 50 years. A considerable portion of our analysis

indicates that there is no trend in extreme wind events for the area.

5.9.1 WBF, YJT and AES40 wind data

The climatology of historical wind data from Stephenville and Burgeo and the AES40

model derived climatological data presented does not indicate a clear trend in the

historical mean or extreme wind conditions of southwest Newfoundland and the Cabot

Strait.

• The yearly average wind observed at Burgeo has a statistically significant

decreasing trend. However, between 1986 and 1990 the annual percentage of

strong winds significantly decreases and there is an equally dramatic increase in

the percentage of less than strong winds. It appears that a significant and

permanent inhomogeneity may have been introduced into the data during that

period that has not been documented.

• The increasing trend in the maximum yearly wind at Burgeo is not statistically

significant.

• Stephenville’s trends are increasing for yearly average and maximum winds but

they are not statistically significant at the 95% confidence level.

• The AES40 data has a statistically significant increasing trend in yearly average

winds. However the increasing trend in maximum winds is not statistically

110

significant at the 95% confidence level when considering the entire data set.

When the data pre-1966 is removed, i.e. the time period being analyzed is the

same as Stephenville and Burgeo, the statistical significance is lost and the overall

trend is similar to Stephenville’s.

• When the duration and intensity of individual wind events for Burgeo and

Stephenville are considered, by analyzing the TII, the trends emulate the trends in

the mean winds. Burgeo’s TII decreasing over time and Stephenville with no

significant trend. However, when the 2 extreme years are removed from the early

period of Burgeo’s time series for comparison, the statistical significance is lost

and it gives the same answer as the Stephenville TII analysis.

5.9.2 AES40 wave data

The results for the wave data emulate the wind data such that the maximum data, yearly

peak maximum wave height and the annual count of significant wave heights greater than

4 metres are not statistically significant. However the mean annual significant wave

height, like the mean average wind, does show a slight, statistically significant, increasing

trend

5.9.3 Wreckhouse data

The data show considerable inter-annual and inter-decadal variability in the frequency of

Wreckhouse events and the maximum sustained wind expected. However, there are no

long term trends present in the data.

5.9.4 100 year tropical storm climatology data

There has been no clear scientifically determined consensus that tropical cyclones are

increasing in frequency or intensity. Active and in-active periods for tropical cyclones in

the Atlantic basin last between 25 and 40 years each and it is possible that we are 10

years into an active cycle (Chris Landsea, http://www.aoml.noaa.gov/hrd/tcfaq/G4.html).

111

The data for the Cabot Strait area that has been analyzed confirms the large inter-annual

and inter-decadal variability in tropical storm frequency with no clear long term trend

present.

5.11 Acknowledgements

We would like to acknowledge the following people who greatly aided in the preparation

of this report:

• Gerard Morin and Rick Fleetwood of the Atlantic Climate Centre in Fredericton,

New Brunswick.

• Neil Burgess, Canadian Wildlife Service

• Jack Cossar, Brian Hulan, Andre Dwyer and Gerry Rockwood of the

Environmental Monitoring Division, Atlantic Region.

• Jerry Pulchan, Environmental Protection

• Dr. Leonard Lye, Memorial University of Newfoundland.

• Doug Mercer, George Parkes, Ming Szeto, Bruce Whiffen and Bridget Thomas of

Environment Canada, Atlantic Region.

• Peter Bowyer, Canadian Hurricane Centre

• Kevin Pike, Environmental Protection

112

6. The Value of Economic Loss Associated with Weather-Related Delays on the

North Sydney-Port aux Basques Marine Transportation Link in 2003 and 2004

Wade Locke

6.1 Introduction

This report was prepared as part of the study entitled: Storm and Wind Impacts on

Transportation, Southwest Newfoundland. The purpose of this report is to estimate the

economic loss associated with weather-related delays associated with Marine Atlantic’s

marine transportation service between North Sydney, Nova Scotia and Port aux Basques,

Newfoundland and Labrador.

The economic costs associated with weather related delays on the North Sydney-Port aux

Basques run averaged $3.5 million per year. The estimates for each year were:

• $2.1 million in 2003; and

• $4.9 million in 2004

6.2 Methodology and Results

There are no published data available on the number and duration of weather-related

delays or on the number of passengers or commercial vehicles affected by these delays

for the marine transportation link between North Sydney, Nova Scotia and Port aux

Basques, Newfoundland and Labrador. However, Marine Atlantic records a daily

Situation Report for the ships that it operates on this run.2 For the purpose of this study,

Marine Atlantic provided the 2003 and 2004 Situation Reports for both the Port aux

Basques and North Sydney ports. This meant that 1,462 reports were reviewed and

analyzed for this study3.

2 The four ships that provided marine transportation services between Port aux Basques and North Sydney between 2003 and 2004 were: the MV Caribou, the MV Joseph and Clara Smallwood, the MV Leif Ericson and the MV Atlantic Freighter. 3 All data are available from Norm Catto on request.

113

These reports indicated for each day of the year and for all of the ships involved in this

route the following information:

• the number of passengers;

• the number of passenger-related vehicles;

• the number of commercial vehicles;

• the time of departure and the time of arrival; and

• whether the ship was late leaving or arriving and the reason why, including

specifying weather-related delays.

For each of the ships operating on the Nova Scotia/Newfoundland and Labrador run, it is

possible to determine the number of weather-related delays and their duration. In

addition, it is possible to determine the number of passengers and commercial traffic that

was affected by these delays. Because of the normal variation in sailing times, only

delays of two hours or more are included in this analysis.

For each weather-related delay of two hour or more, the corresponding number of

passengers and commercial vehicles are recorded. As well for each delay, the number of

passenger hours are calculated as the product of the number of passengers and the

number of hours involved in that specific delay. Similarly, the number of commercial

vehicle hours is calculated as the product of the number of commercial vehicles involved

in the delay times the number of hours associated with the delay. This information is

presented in Table 6.1.

114

Table 6.1: Weather Related Delays on the North Sydney-Port aux Basques Run

for Marine Atlantic in 2003 and 2004

Ship Delay Hours

Passengers Commercial

Vehicles Passengers

Hours

Commercial Vehicles

Hours

Year - 2003

Caribou 168 2,414 918 25,388 9,706

Smallwood 257 2,266 730 35,733 14,303

Atlantic Freighter 3 6 61 18 183

Leif Ericson 232 1,881 946 22,298 12,099

Combined 660 6,567 2,655 83,437 36,291

Year - 2004

Caribou 417 4,061 1,251 83,665 22,244

Smallwood 461 4,185 1,724 70,177 25,806

Atlantic Freighter 150 60 421 1,106 7,072

Leif Ericson 349 2,300 862 50,818 18,955

Combined 1,377 10,606 4,258 205,766 74,077

Average – 2003 and 2004

Caribou 292 3,238 1,085 54,526 15,975

Smallwood 359 3,226 1,227 52,955 20,055

Atlantic Freighter 77 33 241 562 3,628

Leif Ericson 291 2,091 904 36,558 15,527

Combined 1,018 8,587 3,457 144,601 55,184

The next step involved trying to assign a value to the lost time associated with these

weather-related delays. In this analysis, it is assumed that the economic value of a

specific unit of time saved from transportation projects is equivalent to, but of opposite

sign, for the same-sized unit of time lost due to weather-related delays.

Transport Canada, in their Guide to Benefit-Cost Analysis4 suggests using $24.00 per

hour in 1990 dollars for the value of business travel by automobile saved from

transportation projects in Canada and using $9.10 per hour in 1990 dollars where

“reliable information on trip purpose (i.e., business/non-business) is not available.” Since

there is no data on which of the passengers are traveling for business or other purposes,

the $9.10 per is used to value the lost time for passengers caused by the weather delay.

The lost time associated with commercial vehicles is valued at $24.00 per hour. In order

to take into account the level of inflation that has occurred between 1990 and 2003 and

4 See http://www.tc.gc.ca/Finance/BCA/en/Section7.htm.

115

2004, these dollar estimates are adjusted to their 2003 and 2004 equivalents utilizing the

Consumer Price Index for Canada (1992=100).5 Making these adjustments yields:

• $11.93 per hour for passengers and $31.46 for commercial vehicles in 2003; and

• $12.15 per hour for passengers and $32.05 for commercial vehicles in 2005.

Applying these values to the weather-related delays generates the economic value of the

lost time on the North Sydney-Port aux Basques run as a result of weather in 2003 and

2004. These estimates are presented in the following table.

Table 6.2: Value of Weather Related Delays on the North Sydney-Port aux Basques

Run for Marine Atlantic in 2003 and 2004

Ship Passengers

Hours

Commercial Vehicles

Hours

Dollar Value of

Passengers Delays

Dollar Value of

Commercial Vehicles Delays

Total Value of Weather Delays

Year – 2003

Caribou 25,388 9,706 $302,879 $305,351 $608,230

Smallwood 35,733 14,303 $426,295 $449,972 $876,267

Atlantic Freighter 18 183 $215 $5,757 $5,972

Leif Ericson 22,298 12,099 $266,015 $380,635 $646,650

Combined 83,437 36,291 $995,403 $1,141,715 $2,137,118

Year –2004

Caribou 83,665 22,244 $1,016,524 $712,904 $1,729,428

Smallwood 70,177 25,806 $852,651 $827,082 $1,679,733

Atlantic Freighter 1,106 7,072 $13,438 $226,658 $240,096

Leif Ericson 50,818 18,955 $617,439 $607,508 $1,224,946

Combined 205,766 74,077 $2,500,051 $2,374,152 $4,874,203

Average – 2003 and 2004

Caribou 54,526 15,975 $659,701 $509,127 $1,168,829

Smallwood 52,955 20,055 $639,473 $638,527 $1,278,000

Atlantic Freighter 562 3,628 $6,826 $116,207 $123,034

Leif Ericson 36,558 15,527 $441,727 $494,071 $935,798

Combined 144,601 55,184 $1,747,727 $1,757,933 $3,505,660

5 The CPI (1992=100) was 93.3 in 1990, 122.3 in 2003 and 124.6 in 2004.

116

6.3 Conclusion

Weather-related delays on the North Sydney-Port aux Basques run had an average

economic loss of $3.5 million per year. The corresponding losses for 2003 and 2004

were $2.1 million and $4.9 million, respectively. These weather-related delays involved

6,567 passengers and 2,655 commercial vehicles in 2003. Weather-related delays were

more frequent in 2004 than in 2003. Specifically, in 2004 there were 10,606 passengers

and 4,258 commercial vehicles that experienced delays on the Port aux Basques-North

Sydney run. Moreover, the number of hours of weather related delays in 2004 were

nearly double those of 2003: 1,377 hours in 2004 versus 660 hours in 2003.

117

7. Summary and Discussion

The southwest coast of Newfoundland has been subjected to many significant

storms resulting in property destruction. The storm of January 2000 is the most recent

event causing significant damage to Channel-Port aux Basques. Analysis of this event

has served as a model for predicting the impact of future storms. The coastline of

southwestern Newfoundland has been subject to at least 5 storm surge events since

January 2000, including those resulting from Hurricane Gustav (12 September 2002) and

Hurricane Frances (2 September 2004).

The estimated economic damage from the January 2000 event exceeded $1

million. However, study of other flooding events in Newfoundland indicates that actual

costs were more than twice the reported figures in all instances. Applying a similar

standard to Port aux Basques would suggest that the actual financial impact of the 21-22

January 2000 storm surge event would approach $ 3 million. This figure does not

account for any losses resulting in disruption to transportation between Port aux Basques

and the remainder of Newfoundland.

The effectiveness of any particular storm at a location depends upon the angle of

wave attack, the number of previous events during the season, and other local factors.

Adjacent beaches can exhibit very different responses to a particular storm, as was

evident on southwestern Newfoundland beaches impacted by Hurricanes Gustav (2002)

and Frances (2004).

Although the relationship between changes in hurricane frequency and magnitude,

and increases in air temperature or sea surface temperature, is not clear at present, and

consensus does not exist, it is apparent that the North Atlantic is currently undergoing a

period of increased hurricane activity. Planning emergency response measures to cope

with the impact of storm surges associated with either hurricanes or mid-latitude cyclones

is necessary. Exercises such as “Operation Dolphin”, conducted at Channel- Port aux

Basques in June 2006 (Seguin et al., 2006) are vital planning and preparatory tools.

Tsunamis are the products of seismic activity, and are not related to climate.

However, their potential impact on the southwest Newfoundland coastline is similar to

that for storm surges. The frequency of tsunami events cannot be assessed at present.

118

Sea level is currently rising at Port aux Basques at ca. 3.3 mm/a. Rising sea level

will allow successive storm surges to rise higher and penetrate further inland. The

shoreline at Port aux Basques Harbour is not sensitive to coastal erosion resulting from

sea level rise. Coastal erosion is occurring along dune-backed sandy shorelines, such as

those at JT Cheeseman and Sandbanks Provincial Parks.

There is no clear net change in tidal regime for the period of tidal records for Port

aux Basques (1935-present). There is, however, an indication of the de-tided residual

increasing over time (approximately 10 cm from 1959 to 2005). Unfortunately, this

increase cannot be accurately interpreted since the data set is highly fragmented, but

some of the increasing trend is likely due to relative sea level rise in the area.

There is a clear seasonal trend in mean water level. The mean and variance is

higher during the fall and winter than in the spring and summer. The atmospheric

pressure exhibits a clear seasonal trend. The pressure is lower with a higher variability

during the late fall and winter.

There is no obvious correlation between the NAO and the tidal constituents, or

storm surge activity (neither by frequency in terms of number of events or duration nor

by maximum surge height). In contrast to the situation in eastern Newfoundland,

considering the North Atlantic Oscillation is not an effective predictor or explanation for

surge activity at Port aux Basques. The annual number of events that exceed 60 cm shows

no real trend and there has been little change in the annual maximum storm surge

elevation.

The maximum water level for extreme events, such as the January 2000 storm,

that have been excluded from the data record can be roughly estimated using basic

statistical correlations based on the recorded meteorological conditions along with the

records of other previous storm/surge events. However, the lack of tide gauge data from

the period surrounding the January 2000 storm represents a serious limitation on

interpretation of this event, and on the ability to recognize long-term trends and predict

the pattern of future storm surge events. Tidal gauge networks require adequate

maintenance and associated funding, in order for significant and productive research to

continue and for public safety.

119

Historically, Wreckhouse Events have been known and respected since the initial

railway construction period. At present, it is not possible to assess if any change has

occurred in the frequency of Wreckhouse Events, as judged from resultant delays to

transportation, over the historical record. The available archival data appears to be

inadequate for such an assessment.

The data show considerable inter-annual and inter-decadal variability in the

frequency of Wreckhouse events and the maximum sustained wind expected.

Differences in analysis and treatment of the same set of meteorological data can produce

seemingly different results. The variations in results reflect the relatively short length of

the data set available, and the difficulties in assessing the data.

Figures 7.1 and 7.2 illustrate data generated during preliminary analysis, and

presented in 2005. The data include the percentage of Wreckhouse winds in excess of 57

km/h, a critical value for the interruption of transportation (Figure 7.1), and the inferred

peak velocity for Wreckhouse winds (Figure 7.2). Both data sets suggest a trend of

increasing Wreckhouse events, based on analysis of data from ca. 1967 to 2004.

Figure 7.1 Percentage of Wreckhouse easterly winds in excess of 57 km/h. Graph

prepared by Dale Foote and Dermot Kearney from Environment Canada data, 2005.

% of WZB easterly winds >= 57 kph

0

5

10

15

20

25

30

1960 1970 1980 1990 2000 2010

Year

%

120

Figure 7.2 Inferred yearly peak Wreckhouse wind velocities, based on data from

ca. 1967 to 2004. Graph prepared by Dale Foote and Dermot Kearney from Environment

Canada data, 2005.

However, subsequent analysis, reported in chapter 5 of this report and including

data from 2005, suggests that the activity of Wreckhouse winds may be much more

variable. Figures 7.3 and 7.4 suggest that there is no apparent long-term trend in the

Wreckhouse wind data, with peak wind velocities (Figure 7.3) and total yearly hours

(Figure 7.4)

Inferred yearly Wreckhouse Peak wind

80

90

100

110

120

130

140

1960 1970 1980 1990 2000 2010

Year

2 m

inu

te P

eak w

ind

(kp

h)

121

C4

xwr wind

2002-04

1999-01

1996-98

1993-95

1990-92

1987-89

1984-86

1981-83

1978-80

1975-77

1972-74

1969-71

1966-68

140

130

120

110

100

90

80

Inferred Wreckhouse Peak Wind

Figure 7.3 Inferred Wreckhouse Peak wind velocities, three year averages, 1966-2004.

Wreckhouse Event Total Yearly Hours (1966, 1997 &1998 Removed)

0

100

200

300

400

500

1960 1970 1980 1990 2000 2010

Year

Ho

urs

Figure 7.4 Wreckhouse Event total yearly hours

122

The available data thus are subject to different interpretations. The increased wind

velocities documented at Port aux Basques, as discussed in chapter 5, indicate that wind

velocities at Wreckhouse would also increase, as these two phenomena are linked

meteorologically. However, increases in mean wind velocity at Port aux Basques may

not translate directly into increases in peak wind velocity at Wreckhouse.

At present, it is prudent to conclude that Wreckhouse events will not become less

frequent in the future, and could become more frequent. The problems of data analysis

also strongly indicate that more detailed, locally specific investigations are necessary.

They further highlight the absence of reliable (or any) archival data, which could be used

to assess a longer-term change in the number and severity of events (replacing a 40-year

recording period, with gaps, with a longer time series). The problem is similar to the

difficulties in interpreting the tidal and storm surge record, which is also relatively short

and contains significant gaps.

Assessment of climate change and variation is contingent on reliable, complete data

recorded over long periods of time. Unfortunately, missing data severely hinders

interpretation of climate patterns for both Port aux Basques and Wreckhouse.

The most recent cost estimates of disruption to highway traffic, summarized by

ADI Limited (1994), suggest that accidents cost ca. $236,000 annually, and delays cost a

further $1.9 million (as measured in 2006 dollars). Thus, the cost of Wreckhouse Events

to the provincial economy is in excess of $ 2.1 million annually. This figure only

includes insured losses and delays, and does not include delays or damage to smaller,

non-commercial vehicles, or costs involved with response to accidents or road

maintenance and repair.

Analysis of trends of mean wind speed indicate a statistically significant increasing

trend at Port aux Basques, combined with statistically significant decreasing trends for

Burgeo and Sydney Airport. These differences reflect the influence of local topography

on wind velocity. They also indicate a necessity for detailed local measurements at

several sites, in order both to establish long-term regional trends and for local forecasting

in aid of marine navigation.

123

There is high interannual variability in westerly wind disruption to marine

transportation. Although easterly winds are of concern for the Wreckhouse area, it is the

westerlies that have the greatest potential to produce stormy conditions in Cabot Strait.

Weather-related delays on the North Sydney-Port aux Basques run had an average

economic loss of $3.5 million per year. The corresponding losses for 2003 and 2004

were $2.1 million and $4.9 million, respectively. These weather-related delays involved

6,567 passengers and 2,655 commercial vehicles in 2003. Weather-related delays were

more frequent in 2004 than in 2003. Specifically, in 2004 there were 10,606 passengers

and 4,258 commercial vehicles that experienced delays on the Port aux Basques-North

Sydney run. Moreover, the number of hours of weather related delays in 2004 were

nearly double those of 2003: 1,377 hours in 2004 versus 660 hours in 2003.

Adding the costs for transportation delays related to storm and wind activity along

the route from Channel- Port aux Basques through Wreckhouse suggests an average loss

of $ 5.6 million annually. These costs represent the current losses due to storms, surges,

and winds, and do not consider the potential of increased future losses due either to

changes in the condition of the infrastructure or to any changes in climate. The available

data suggest that the climate regime, as it currently exists, is sufficient to have significant

economic impact to transportation in Southwestern Newfoundland.

124

8. References

ADI Limited, 1994. Wreckhouse Wind Information System. Report to Government of

Newfoundland, Dept. of Works, Services and transportation, 5 August 1994, Project

3596-001.1. 41 pp.

Anderson, T.W., Macpherson, J.B., 1994. Wisconsin Late-glacial environmental change

in Newfoundland: A regional review. Journal of Quaternary Science 9, 171-178.

Bacon, S.J., and Carter, D.J.T., 1991. Wave climate changes in the North Atlantic and

the North Sea. International Journal of Climatology 11, 545-588.

Banfield, C., 1981. The Climatic Environment of Newfoundland. In Macpherson, A.G.

and Macpherson, J.B., eds., The Natural Environment of Newfoundland Past and Present.

Memorial University, St. John's, p. 83-153.

Banfield, C. E., 1993. Newfoundland Climate: Past and Present. In Robertson, A.,

Porter, S., and Brodie, G., eds. Climate and Weather of Newfoundland. St. John's,

Creative Publishing, 13-32.

Bateman, E., 2000. High winds, seas wreak devastation. The Gulf News, Monday,

January 31, 2000: Port-aux-Basques.

Bostock, H., 1970. Physiography of Canada. In Douglas, R.J.W., ed. Geology of

Canada, Geological Survey of Canada, Economic Geology Report 1, 3-29.

Bruce, J., 2002. Consequence Analysis of Storm Surge in the Charlottetown, Prince

Edward Island area. Ottawa: Office of Critical Infrastructure and Emergency

Preparedness, Government of Canada.

Canadian Hydrographic Service, 2004. Canadian Tide and Current Tables: Atlantic

Coast and Bay of Fundy. Department of Fisheries and Oceans Canada: Ottawa.

Canadian Hydrographic Service, 2005. Canadian Tide, Current and Water Levels.

Department of Fisheries and Oceans Canada. http://www.lau.chs-shc.dfo-mpo.gc.ca.

Cardone, VJ., and Swail, VR., 1995. Uncertainty in Prediction of Extreme Storm Seas.

Proceedings 4th International Workshop on Wave Hindcasting and Forecasting, October

16-20, 1995, Banff, Alberta. Environment Canada, 1-20.

Cardone, VJ, Jensen, RE, Resio DT, Swail VR, and Cox AT, 1995. Evaluation of

Contemporary Ocean Wave Models in Rare Extreme Events: The Halloween Storm of

125

October 1991 and the Storm of the Century of March 1993. Journal of Atmospheric and

Ocean Technology, 13, 198-230.

Catto, N.R., 1994. Anthropogenic Pressures and the dunal coasts of Newfoundland. In

Wells PG and Ricketts PJ, eds., Coastal Zone Canada 1994, Co-operation in the Coastal

Zone, Bedford Institute of Oceanography, 5, 2266-2286.

Catto, N.R., 2002a. Anthropogenic Pressures on Coastal Dunes, southwest

Newfoundland. The Canadian Geographer, 46, 17-32.

Catto, N.R., 2002b. Geomorphology of Dune-backed Coasts, southwest Newfoundland,

and Implications for Piping Plover Habitat. Report to Habitat Stewardship Initiative,

Marine and Mountain Economic Development Board, and Tourism, Culture and

Recreation, Newfoundland and Labrador.

Catto, N.R., 2004. Impacts of Recent Climatic Variability on Tourism, Gulf of St.

Lawrence Coastline. Coastal Zone Canada 2004, St. John’s, June 2004.

Catto, N.R. 2006a. Impacts of Climate Change and Variation on the Natural Areas of

Newfoundland and Labrador. Report, Ministry of the Environment, Newfoundland and

Labrador, 160 pp.

Catto, N.R., 2006b. Natural Hazard and Vulnerability Assessment in Atlantic Canada:

Review, Progress, and Challenges. McGill-Queen’s University Press, in press

Catto, N.R., 2006c. More than 16 Years, More than 16 Stressors: Evolution of a

Reflective Gravel Beach 1989-2005. Géographie physique et Quaternaire, in press.

Catto, N.R., and Hickman, H., 2004. Flood Sensitivity Analysis for Newfoundland

Communities. Ottawa: Office of Critical Infrastructure Protection and Emergency

Preparedness Canada.

Catto, N.R., Scruton, D., and Ollerhead, N., 2003. The Coastline of Eastern

Newfoundland. Canadian Technical Report of Fisheries and Aquatic Sciences, 2495.

Ottawa: Fisheries and Oceans Canada.

Chmura, G., Helmer, L.L., Beecher, C.B., and Sunderland, E.M., 2001. Historical rates of

salt marsh accretion on the outer Bay of Fundy. Canadian Journal of Earth Sciences, 38,

1081-1092

Cote, P. W., 1989. Ice limits, eastern Canadian seaboard. Environment Canada,

Atmospheric Environment Service, Ice Centre, Climatology and Applications.

126

Daigle, R., Forbes, D., Parkes, G., Ritchie, H., Webster, T., Bérubé, D., Hanson, A.,

DeBaie, L., Nichols, S., and Vasseur, L., 2006. Impacts of sea level rise and climate

change on the coastal zone of southeastern New Brunswick. Environment Canada report.

510pp (in press)

Damman, W.H., 1983. An ecological subdivision of the island of Newfoundland. W.

Junk, The Hague, 648 p.

Davis, A.M., 1993: The Initiation and Development of Peatlands in Newfoundland and

their response to Global Warming. In Hall, J., and Wadleigh, M., eds., The Scientific

Challenge of our changing environment, Royal Society of Canada, IR93-2, 24-25

Davis, A. M., Wallace, B., McAndrews, J., 1987, Palaeoenvironment and the

Archaeological Record at the L'Anse au Meadows Site, Northern Newfoundland.

INQUA, 12th Conference Abstracts, Ottawa, p. 153.

Dawson, A.G., 1999. Linking tsunami deposits, submarine slides, and offshore

earthquakes. Quaternary International, 60, 119-126.

Debernard, J., Sætra, O., and Røed, R.P. 2002. Future wind, wave and storm surge

climate in the northern North Atlantic. Climate Research 23, 39-49

Desjardins, S., 1995. Numerical Simulation by the MC2 of a Downslope Windstorm

Event: The Suete Wind. Technical Note MWC 95-001.

Dickson, W. L., Delaney, P.W., Poole, J., and Blackwood, R.F., 1996. Ramea, Burgeo

and La Poile District, Newfoundland. Government of Newfoundland and Labrador,

Department of Natural Resources, Geological Survey, Map 94-221.

Doherty, F., 1991. Coastal Easterlies in Stable Southerly Flow Along the South Coast of

Newfoundland. Environment Canada, Atlantic Region Technical Notes 91-003(N).

Emanuel, K., 2005. Increasing destructiveness of tropical cyclones over the past 30 years.

Nature 436, 686-688.

Environment Canada, 2004. Hurricane Juan. Retrieved April 26, 2006, from

http://www.atl.ec.gc.ca/weather/hurricane/juan/

Environment Canada, 2005. Canadian Climate Normals: 1971-2000.

http://climate.weatheroffice.ec.gc.ca/climate_normals/results_e.html.

Etter, D.M., 1993. Engineering problem solving with MATLAB. Prentice Hall: New

Jersey, USA.

127

Fenco Newfoundland Limited, 1985. Hydrotechnical study of the Badger and Rushy

Pond areas. Report to Newfoundland & Labrador Department of Environment, and

Environment Canada.

Fine, I.V. Rabinovichb, A.B., Bornholdd, B.D., Thomsonb, R.E., and Kulikov E.A.

2005. The Grand Banks landslide-generated tsunami of November 18, 1929: preliminary

analysis and numerical modeling. Marine Geology, 215, 45–57

Forbes, D.L., Parkes, G.S., Manson, G.K., Ketch, L.A., 2004. Storms and shoreline

retreat in the southern Gulf of St. Lawrence. Marine Geology, 210: 169-204.

Forbes, D., Parkes, G., O’Reilly, C., Daigle, R., Taylor, R., Catto, N., 2000. Storm surge,

sea-ice, and wave impacts of the 21-22 January 2000 storm in coastal communities of

Atlantic Canada. Canadian Meteorological and Oceanographic Society, 34th Congress,

Victoria, BC, 29 May – 2 June 2000.

Goldenberg, SB, Shapiro LJ, Landsea, CW, 1997. The Hyper-active 1995 Atlantic

Hurricane Season: A Spike or a Harbinger of Things to Come? In Shaw, R.W., ed.

Climate Change and Climate Variability in Atlantic Canada. Dartmouth: Environment

Canada, Atlantic Region, Occasional paper 9, 113-119.

Goldenberg, SB, Landsea, CW, Mestas-Nuñez, AM, and Gray, WM, 2001. The Recent

Increase in Atlantic Hurricane Activity: Causes and Implications. Science, 293, 474-479.

Gornitz, V., 1991. Global coastal hazards from future sea level rise. Palaeogeography,

palaeoclimatology, palaeoecology, 89, 379-398.

Gornitz, V., White TW, Cushman RM 1991. Vulnerability of the US to future sea-level

rise. Coastal Zone ‘91, Proceedings of the 7th Symposium on Coastal and Ocean

Management, American Society of Civil Engineers, 1345-1359.

Gornitz, V., Daniels RC, White TW, and Birdwell, KR 1993. The development of a

coastal risk assessment database: Vulnerability to sea-level rise in the US Southeast.

Washington: US Government Report DE-AC05-84, Environmental Sciences Division

Publication 3999.

Government of Newfoundland and Labrador, 2005. Population projections – Total

province by economic zone. Retrieved April 25, 2006, from

http://www.economics.gov.nf.ca/population/byzone.asp

128

Harding, L., 1992. St. John’s to Port aux Basques: When the Going Was Bad. The Beaver

27, n. 1, 34-36

H. John Heinz Center for Science, Economics, and the Environment, 2000. Developing

and evaluating mitigation strategies. In H. John Heinz Center for Science, Economics,

and the Environment, The Hidden Costs of Coastal Hazards. Island Press, Washington,

139-168.

Hickman, H., 2006. Flood Sensitivity and Vulnerability of Newfoundland Communities.

M. Environmental Science thesis, Memorial University.

Hurrell, J.W., Kushnir, Y., Ottersen, G., Visbeck, M., 2003. An overview of the North

Atlantic Oscillation. Geophysical Monograph, 134: 1-35.

Ingram, D., 2004. Coastal Geomorphology, Erosion, and Anthropogenic Stresses,

Sandbanks Provincial Park, Southwestern Newfoundland. B.Sc. thesis, Department of

Geography, Memorial University.

Ingram, D., 2005. An Investigation of the role of Tidal Variation on Storm Surge

Elevation and Frequency in Port-aux-Basques, Newfoundland. Research Report, Master

of Environmental Science, Memorial University.

Ingram, D., DeYoung, B., 2005. The analysis of Tides in Port-aux-Basques,

Newfoundland: 1935-2005. Physics and Physical Oceanography Data Report 2005-1,

Department of Physics and Physical Oceanography, Memorial University.

Karn, J., 2004. Atlantic Canada wind trends and variability: 1953 – 2003. Research

Report, Master of Environmental Science. Memorial University.

Knutsen, T.R., and Tuyela, R.E., 2004. Impact of CO2-Induced Warming on Simulated

Hurricane Intensity and Precipitation: Sensitivity to the Choice of Climate Model and

Convective Parameterization. Journal of Climate, 17, 3477–3495.

Kreibich, H., Müller, M., Thieken, A., Axer, T., and Merz, B. 2004. Flood losses in

private households – lessons learned from the Elbe and Danube floods in August 2002.

Geophysical Research Abstracts, 6(00999).

Kumar, A., Burton, I., and Etkin, D. 2001. Managing flood hazard and risk: Report of an

independent expert panel. Office of Critical Infrastructure Protection and Emergency

Preparedness (OCIPEP).

129

Landsea, C.W., Pielke, R.A., Mestas-Nuñez, A.M., and Knaff, J.A., 1999. Atlantic Basin

Hurricanes: Indices of Climatic Changes. Climatic Change, 42, 89-129.

Large, W.G. and Pond, S., 1981. Open ocean momentum flux measurements in moderate

to strong winds. Journal of Physical Oceanography, 11: 324-336.

LeBlond, P.H., and Mysak, L.A., 1978. Waves in the Ocean. Elsevier Oceanography

Series, 20. Elsevier: New York.

Lewis, PJ, 1997. Climate Trends in Atlantic Canada. In Shaw, R.W., ed. Climate Change

and Climate Variability in Atlantic Canada. Dartmouth: Environment Canada, Atlantic

Region, Occasional paper 9, 180-183.

Lewis, P.J. and Moran, M.D. 1984. Severe storms off Canada’s east coast catalogue

summary for the period 1957-1983. The Canadian Climate Program. Concord Scientific

Corporation.

Lingard, M., 1997. Next Stop: Wreckhouse: More Chats, Stats, and Snaps of the

Newfoundland Railway. Grand Falls-Windsor: Mont Lingard Publishing.

Lines, G., Pancura, M., Landeer, C., 2003. Building climate change scenarios of

temperature and precipitation in Atlantic Canada using the statistical downscaling model

(SDSM). 14th Symposium on Global Change and Climate Variations, American

Meteorological Society Annual Meeting, Long Beach CA, 1-25

Liverman, D.G.E., Batterson, M.J., Taylor, D., and Ryan, J., 2001. Geological hazards

and disasters in Newfoundland. Canadian Geotechnical Journal, 38, 936-956.

Macpherson, J.B., 1995. A 6 ka reconstruction for the island of Newfoundland from a

synthesis of Holocene lake-sediment pollen records. Géographie physique et

Quaternaire, 49, 163-182.

Marine Atlantic, 2003. Marine Atlantic Incorporated website.

http://www.marineatatlantic.ca

McCuaig, S, Bell. T., 2005. Is there a record of Tsunami deposits in Newfoundland? A

Case Study from the Burin Peninsula. Abstract, Geological Association of Canada

Annual Meeting, Halifax, NS.

McCulloch, M M; Forbes, D L; Shaw, R D., 2002. Coastal Impacts of Climate Change

and Sea-Level Rise on Prince Edward Island. Ottawa: Geological Survey of Canada,

Open File 4261.

130

Milne, J. 2002. Better flood management: the insurance perspective. Proceedings of the

Institution of Civil Engineers, 151(4), 283-286.

Morris, D., 1977. We met the devil and we beat him. The Rounder, 2 (5), January 1977,

10.

Municipal and Provincial Affairs, 2000. Funding approved to address storm damages.

Press release, May 23, 2000. Government of Newfoundland and Labrador, St. John’s.

Murty, T.S., 1984. Storm surges – meteorological ocean tides. Canadian Bulletin of

Fisheries and Aquatic Sciences, 212. Fisheries and Oceans Canada, Ottawa.

Neu, HJA., 1982. 11-year deep water wave climate of Canadian Atlantic waters.

Fisheries and Oceans Canada, Canadian Technical Report of Hydrography and Ocean

Sciences, 13.

Parkes, G., Forbes, D.L. and Ketch, L.A., 2006. Sea-level rise and regional subsidence in

southeastern New Brunswick. In Daigle et al. Impacts of sea level rise and climate

change on the coastal zone of southeastern New Brunswick. (CCAF –Environment

Canada report in preparation).

Parkes, G.S., Ketch, L.A., 2002. Storm-surge climatology. Geological Survey of

Canada, Open File 4261.

Parson, E.A, Corell, R. T. W., Barron, E.J., Burkett, V., Janetos, A., Joyce, L. Karl, T.R,

MaCcracken, M.C., Melillo, J., Morgan, M.G., Schimel, D.S., and Wilbanks, T. 2003.

Understanding climatic impacts, vulnerabilities, and adaptation in the United States:

building a capacity for assessment. Climatic Change, 57, 9–42.

Pielke, R.A., Landsea, C., Mayfield, M., Laver, J., and Pasch, R., 2005. Hurricanes and

Global Warming. Bulletin of the American Meteorological Society, 86, 1571-1575.

Pittman, D., and Catto, N.R., 2001. Newfoundland’s Coastal Dunes as Geoindicators. In

A.R. Berger and D.G. Liverman (Eds.), Geoindicators for Ecosystem monitoring in Parks

and Protected Areas (Parks Canada Ecosystem Science Review Reports 018, 43-51).

Pugh, D.T., 1987. Tides, Surges, and Mean Sea-level. Wiley: New York.

Riley, G.C., 1959. Geology, Burgeo-Ramea, Newfoundlan’. Geological Survey of

Canada, Map 22-1959.

Ruffman, A., 1991. The 1929 ‘Grand Banks’ Earthquake and the historical record of

Earthquakes and Tsunamis in Eastern Canada. Proceedings, Geological Survey of

131

Canada Workshop on Eastern Seismicity Source Zones for the 1995 Seismic Hazard

Maps, March 18-19. Ottawa, Geological Survey of Canada Open File 2437, 193.

Ruffman, A., 1993. Reconnaissance Search on the South Coast of the Burin Peninsula,

Newfoundland, for tsunami-laid sediments deposited by the 'tidal wave' following the

November 18, 1929 Laurentian Slope Earthquake, August 17 - September 2, l993.

Geomarine Associates Ltd., Halifax, Nova Scotia, Project 90-19, Contract Report for

Seismology, Lamont-Doherty Earth Observatory of Columbia University, Palisades, New

York, Contract No. NRC-04-92-088, U.S. Nuclear Regulatory Commission, Washington,

D.C., September 26, 228 pp.

Ruffman, A., 1995a. Tsunami Runup Maps as an Emergency Preparedness Planning

Tool: the November 18, 1929 tsunami in St. Lawrence, Newfoundland, as a case study.

Geomarine Associates Ltd., Halifax, Nova Scotia, Project 94-14, for Emergency

Preparedness Canada, 399 p.

Ruffman, A., 1995 b. Comment on: “The Great Newfoundland Storm of 12 September

1775” by Anne E. Stevens and Michael Staveley. Bulletin of the Seismological Society of

America, 85, 646-649.

Ryan, M., 2000. Surveys confirm it: biggest wave since 1929 Burin tidal. The Gulf

News, Monday, February 7, 2000: Port-aux-Basques.

Seguin, J., et al., 2006. Assessment of the Capacity of the Emergency Response and

Public Health Systems in Atlantic Coastal Communities to Cope with and Adapt to

Extreme Weather Events Exacerbated by a Changing Climate. Canadian Climate Impacts

and Adaptations Directorate, report.

Shapiro, L.J., and Goldenberg, S.B., 1998. Atlantic Sea Surface Temperatures and

Tropical Cyclone Formation. Journal of Climate, 11, 578–590.

Shaw, J., 2006. Adaptation to rising sea level in the Bras d´Or Lakes, Canada´s largest

inland sea. Canadian Climate Impacts and Adaptations Directorate, report.

Shaw, J., Taylor, R B, Forbes, D L, Solomon, S, 2001. Sea level rise in Canada. In

Brooks, G R., ed., A synthesis of geological hazards in Canada; Geological Survey of

Canada Bulletin 548, 225-226.

Shaw, J., Taylor, R.B., Forbes, D.L., Solomon, S., and Ruz, M.-H., 1998. Sensitivity of

the coasts of Canada to Sea-level Rise. Geological Survey of Canada, Bulletin 505.

132

Stevens, A.E. 1995. Reply to Comments on “The great Newfoundland storm of 12

September 1775”. Bulletin of the Seismological Society of America, 85, 650-652.

Stevens, A.E., and Staveley, M., 1991. The great Newfoundland storm of 12 September

1775. Bulletin of the Seismological Society of America, 81, 1398-1402.

Sugi, M., Noda, A., and Sato, N., 2002. Influence of the Global Warming on Tropical

Cyclone Climatology: An Experiment with the JMA Global Model. Journal of The

Meteorological Society of Japan 80, 249-272.

Sutherland, C.H., B. Janz, and A.P. Moakler, 1963. Study of Abnormal Winds in

Southwestern Newfoundland and Train Delays 1956-1962 Inclusive. Department of

Transport Meteorological Branch CIR – 3903 TEC – 484

Swail, VR, 1997. Analysis of Climate Variability in Ocean Waves in the Northwest

Atlantic Ocean. In Shaw, R.W., ed., Climate Change and Climate Variability in Atlantic

Canada. Environment Canada, Atlantic region, occasional paper 9, 313-317.

Tapsell, S. and Tunstall, S. 2003. An examination of the health effects of flooding in the

United States. Journal of Meteorology, 28, 341-349.

Trenberth, K.E., 2005. Uncertainty in Hurricanes and Global Warming. Science 308,

1753-1754.

Trenberth, K.E., Dai, A.G., Rasmussen, R.M., Parsons, D.B., 2003. The changing

character of precipitation. Bulletin of the American Meteorological Society 84, 1205-

1217.

TRO Engineering, 2002. TL 214: Condition Assessment and Recommendations for

Upgrading. http://n225h099.pub.nf.ca/hyd2003B/files/applic/AppSectionGtab3.pdf,

Retrieved 20 January 2006.

Van Staal, C.R., Hall, L, Schofield, D., Valverde, P., 1996. Geology, Port aux Basques

area, Newfoundland. Geological Survey of Canada, Open File 3165.

Weather Engineering Corporation of Canada, Ltd. 1982. Study of Wreckhouse Wind

Effect in the Codroy Valley Along TL-214. Report to Newfoundland & Labrador Hydro.

Webster, P.J., Holland, G.J., Curry, J.A. and Chang, H.-R., 2005. Changes in Tropical

Cyclone Number, Duration, and Intensity in a Warming Environment. Science 309,

1844-1846.

133

Windsor, H., 1956. New C.N.R. Terminal at Port-aux-Basques. Atlantic Guardian,

September – October 1956: St. John’s.

134

9. Appendices

Appendix A

Station Histories STEPHENVILLE - 48 32’/58 33’

DATE

ANEMOMETER ELEVATION/

HEIGHT

ANEMOMETER TYPE

WIND RECORDER TYPE

MANNED/ AUTO

WIND SPEED AVERAGING

PERIOD (MINUTES)

Jan/53 Unknown/ 3.96m AN-GMQ-11 UNKNOWN MANNED ? June/66 7.92m/ 6m U2A MONRO MANNED 1 Sept/66 7.92m/ 10.0m U2A MONRO MANNED 1 ?/1980 7.92m/ 10.0m 78D ELECTRONIC MANNED 2

BURGEO- 47 37’/57 37’

DATE

ANEMOMETER ELEVATION/

HEIGHT

ANEMOMETER TYPE

WIND RECORDER TYPE

MANNED/ AUTO

WIND SPEED AVERAGING

PERIOD (MINUTES)

July/66 22.73m/ 10.0m 45B WIND RUN MANNED 60 Sept/69 22.73m/ 10.0m U2A MONRO MANNED 10

Oct 02/91 22.73m/ 10.0m U2A MONRO MANNED and AUTO8

10

July 31/95 22.73m/ 10.0m U2A N/A AUTO8 10

SYDNEY - 46 10’/60 03’

DATE

ANEMOMETER ELEVATION/

HEIGHT

ANEMOMETER TYPE

WIND RECORDER TYPE

MANNED/ AUTO

WIND SPEED AVERAGING

PERIOD (MINUTES)

June/53 55.5m/ 19.5m 45B WIND RUN MANNED 60 August/55 55.5m/ 19.8m 45B WIND RUN MANNED 60

November/59 55.5m/ 19.5m U2A MONRO MANNED ? February/65 55.5m/ 16.4m 45B WIND RUN MANNED 60

July/66 55.5m/ 10.0m U2A MONRO MANNED 1 August/68 55.5m/ 10.0m U2A MONRO MANNED 1 June/72 55.5m/ 10.0m U2A/R MONRO MANNED 1 ?/1980 55.5m/ 10.0m U2A/R MONRO MANNED 2

November/96 55.5m/ 10.0m 78D ELECTRONIC MANNED 2

135

PORT AUX BASQUES

SITE DESCRIPTION

DATE

ANEMOMETER ELEVATION/

HEIGHT

ANEMOMETER TYPE

WIND DISPLAY

WIND SPEED AVERAGING

PERIOD (MINUTES)

Telegraph and Cable Office

1929

45.5m/ 6.1m

?

?

?

Loran Transmitting Station

1955

N/A

N/A

N/A

N/A Mouse Island 47 34’/ 59 10’

June/66 17.0m/ 13m U2A MONRO 10

800m NNE of previous location

August/83 49m/ 10m U2A DIAL 10

Manned/Auto (no change in

location)

October/91 49m/ 10m U2A DIAL 10

AWOS (no change in

location)

Spring/98 49m/ 10m 78D ELECTRONIC 10

ST. ANDREWS-47 46’ /59 19

’

DATE

ANEMOMETER ELEVATION/

HEIGHT

ANEMOMETER TYPE

WIND RECORDER TYPE

MANNED/ AUTO

WIND SPEED AVERAGING

PERIOD (MINUTES)

1953-May 1966 21.0/1.83m 45B NONE MANNED 60

WRECKHOUSE-47 42’ / 59 18

’

DATE

ANEMOMETER ELEVATION/

HEIGHT

ANEMOMETER TYPE

WIND RECORDER TYPE

MANNED/ AUTO

WIND SPEED AVERAGING

PERIOD (MINUTES)

Jan 2000 31.7 m/10m R.M Young NONE AUTO 2

136

Appendix B: Seasonal Graphs

Burgeo Seasonal Wind Speed Categories

DJF 1970-2004

0.01

0.1

1

10

1001

97

0

19

73

19

76

19

79

19

82

19

85

19

88

19

91

19

94

19

97

20

00

20

03

Year

% O

cc

urr

en

ce

% < 19

% Strong

% Gale

% Storm

Burgeo Seasonal Wind Speed Categories

MAM 1970-2004

0.01

0.1

1

10

100

19

70

19

73

19

76

19

79

19

82

19

85

19

88

19

91

19

94

19

97

20

00

20

03

Year

% O

cc

urr

en

ce

% < 19

% Strong

% Gale

% Storm

137

Burgeo Seasonal Wind Speed Categories

JJA 1970-2004

0.01

0.1

1

10

100

19

70

19

73

19

76

19

79

19

82

19

85

19

88

19

91

19

94

19

97

20

00

20

03

Year

% O

cc

urr

en

ce

% < 19

% Strong

% Gale

% Storm

Burgeo Seasonal Wind Speed Categories

SON 1970-2004

0.01

0.1

1

10

100

19

70

19

73

19

76

19

79

19

82

19

85

19

88

19

91

19

94

19

97

20

00

20

03

Year

% O

cc

urr

en

ce

% < 19

% Strong

% Gale

% Storm

138

Port aux Basques Seasonal Wind Speed Categories

DJF 1967-2004 (1998 removed)

0.01

0.1

1

10

100

19

67

19

70

19

73

19

76

19

79

19

82

19

85

19

88

19

91

19

94

19

97

20

00

20

03

Year

% O

cc

urr

en

ce

% < 19

% Strong

% Gale

% Storm

Port aux Basques Seasonal Wind Speed Categories

MAM 1967-2004 (1998 removed)

0.01

0.1

1

10

100

19

67

19

70

19

73

19

76

19

79

19

82

19

85

19

88

19

91

19

94

19

97

20

00

20

03

Year

% O

cc

urr

en

ce

% < 19

% Strong

% Gale

% Storm

139

Port aux Basques Seasonal Wind Speed Categories

JJA 1967-2004

0.01

0.1

1

10

100

19

67

19

70

19

73

19

76

19

79

19

82

19

85

19

88

19

91

19

94

19

97

20

00

20

03

Year

% O

cc

urr

en

ce

% < 19

% Strong

% Gale

% Storm

Port aux Basques Seasonal Wind Speed Categories

SON 1967-2004 (1997 removed)

0.01

0.1

1

10

100

19

67

19

70

19

73

19

76

19

79

19

82

19

85

19

88

19

91

19

94

19

97

20

00

20

03

Year

% O

cc

urr

en

ce

% < 19

% Strong

% Gale

% Storm

140

Stephenville Airport Seasonal Wind Speed Categories

DJF 1967-2004

0.01

0.1

1

10

100

19

67

19

70

19

73

19

76

19

79

19

82

19

85

19

88

19

91

19

94

19

97

20

00

20

03

Year

% O

cc

urr

en

ce

% < 19

% Strong

% Gale

% Storm

Stephenville Airport Seasonal Wind Speed Categories

MAM 1967-2004

0.01

0.1

1

10

100

19

67

19

70

19

73

19

76

19

79

19

82

19

85

19

88

19

91

19

94

19

97

20

00

20

03

Year

% O

cc

urr

en

ce

% < 19

% Strong

% Gale

% Storm

141

Stephenville Airport Seasonal Wind Speed Categories

JJA 1967-2004

0.01

0.1

1

10

100

19

67

19

70

19

73

19

76

19

79

19

82

19

85

19

88

19

91

19

94

19

97

20

00

20

03

Year

% O

cc

urr

en

ce

% < 19

% Strong

% Gale

% Storm

Stephenville Airport Seasonal Wind Speed Categories

SON 1967-2004

0.01

0.1

1

10

100

19

67

19

70

19

73

19

76

19

79

19

82

19

85

19

88

19

91

19

94

19

97

20

00

20

03

Year

% O

cc

urr

en

ce

% < 19

% Strong

% Gale

% Storm

142

Sydney Airport Seasonal Wind Speed Categories

DJF 1969-2004

0.01

0.1

1

10

100

19

69

19

72

19

75

19

78

19

81

19

84

19

87

19

90

19

93

19

96

19

99

20

02

Year

% O

cc

urr

en

ce

% < 19

% Strong

% Gale

% Storm

Sydney Airport Seasonal Wind Speed Categories

MAM 1969-2004

0.01

0.1

1

10

100

19

69

19

72

19

75

19

78

19

81

19

84

19

87

19

90

19

93

19

96

19

99

20

02

Year

% O

cc

urr

en

ce

% < 19

% Strong

% Gale

% Storm

143

Sydney Airport Seasonal Wind Speed Categories

JJA 1969-2004

0.01

0.1

1

10

100

19

69

19

72

19

75

19

78

19

81

19

84

19

87

19

90

19

93

19

96

19

99

20

02

Year

% O

cc

urr

en

ce

% < 19

% Strong

% Gale

% Storm

144

Appendix C: Wind Rose Diagrams

Joint Frequency Distribution

Burgeo, NL (1966-2004) February

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

4.52

5.43 4.01

6.19

6.03

7.84

4.15

2.03

1.48 2.39

2.32

3.73

3.57

12.89

16.65

9.99 6.80

Wind Speed ( Kilometers Per Hour)

1 11 21 31 41 51

Joint Frequency Distribution

Burgeo, NL (1966-2004) March

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

5.23

4.71 4.33

7.43

7.66

11.22

4.96

2.35

1.96 2.97

2.98

4.13

4.71

13.18

9.99

6.41

5.76

Wind Speed ( Kilometers Per Hour)

1 11 21 31 41 51

Joint Frequency Distribution

Burgeo, NL (1966-2004) April

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

5.84

4.26 3.80

6.06

8.56

18.21

7.67

3.08

2.37 2.37

2.43

3.67

4.53

11.79

6.16

4.84

4.36

Wind Speed ( Kilometers Per Hour)

1 11 21 31 41 51

145

Joint Frequency Distribution

Burgeo, NL (1966-2004) May

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

7.58

2.72 2.39

4.51

7.04

21.20

11.06

3.54

2.65 2.28

2.31

4.89

5.18

11.00

4.61

3.77

3.28

Wind Speed ( Kilometers Per Hour)

1 11 21 31 41 51

Joint Frequency Distribution

Burgeo, NL (1966-2004) June

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

9.19

2.15 1.51

2.31

4.48

23.78

12.02

4.34

2.85 2.40

2.73

5.79

6.00

11.05

4.12

3.12

2.17

Wind Speed ( Kilometers Per Hour)

1 11 21 31 41 51

Joint Frequency Distribution

Burgeo, NL (1966-2004) July

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

11.37

1.39 1.19

1.63

4.11

24.49

11.30

4.60

3.27 2.65

2.69

5.54

7.08

10.72

4.05

2.33

1.60

Wind Speed ( Kilometers Per Hour)

1 11 21 31 41 51

Joint Frequency Distribution

Burgeo, NL (1966-2004) August

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

12.19

1.93 1.54

2.34

5.03

16.12

7.05

3.44

3.14 4.08

4.34

6.77

7.85

12.53

5.54

3.57

2.54

Wind Speed ( Kilometers Per Hour)

1 11 21 31 41 51

146

Joint Frequency Distribution

Burgeo, NL (1966-2004) September

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

9.73

3.03 2.38

3.90

5.44

11.77

5.30

2.68

2.91 4.05

3.73

5.65

6.82

15.15

8.10

5.55

3.81

Wind Speed ( Kilometers Per Hour)

1 11 21 31 41 51

Joint Frequency Distribution

Burgeo, NL (1966-2004) October

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

6.68

3.62 3.22

4.27

5.72

9.86

4.40

2.81

2.38 3.69

3.68

4.81

6.54

16.55

9.87

6.90

5.01

Wind Speed ( Kilometers Per Hour)

1 11 21 31 41 51

Joint Frequency Distribution

Burgeo, NL (1966-2004) November

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

4.07

3.60 3.15

5.60

6.99

9.84

3.87

2.22

2.17 3.66

3.04

4.02

4.61

15.63

14.35

7.95

5.24

Wind Speed ( Kilometers Per Hour)

1 11 21 31 41 51

Joint Frequency Distribution

Burgeo, NL (1966-2004) December

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

3.56

5.07 3.99

5.46

6.02

7.72

3.84

2.34

2.02 3.21

2.67

3.25

4.58

14.79

16.28 8.69

6.48

Wind Speed ( Kilometers Per Hour)

1 11 21 31 41 51

147

Joint Frequency Distribution

Port aux Basques, NL (1966-2004) January

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

1.37

4.21 3.16

3.28

5.40

8.17

2.69

2.32

2.33 3.63

3.26

2.77

5.54

15.61

16.03

14.38

5.86

Wind Speed ( kmh)

1 11 21 31 41 51

Joint Frequency Distribution

Port aux Basques, NL (1966-2004) February

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

2.35

4.61 3.18

2.96

6.34

10.09

3.39

2.03

1.72 3.03

2.71

2.28

4.21

17.28

16.56

12.49

4.78

Wind Speed ( kmh)

1 11 21 31 41 51

Joint Frequency Distribution

Port aux Basques, NL (1966-2004) March

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

2.98

4.58 2.93

3.44

7.20

17.55

4.41

2.16

2.06 3.20

1.90

1.60

3.50

14.16

14.00

10.05 4.30

Wind Speed ( kmh)

1 11 21 31 41 51

Joint Frequency Distribution

Port aux Basques, NL (1966-2004) April

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 10% intervals.Calms included at center.

2.77

4.41 2.77

2.55

6.89

28.52

5.48

2.27

1.57 1.90

1.40

1.23

2.68

11.11

13.82

7.15

3.50

Wind Speed ( kmh)

1 11 21 31 41 51

148

Joint Frequency Distribution

Port aux Basques, NL (1966-2004) May

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 10% intervals.Calms included at center.

3.34

3.13 1.67

1.88

5.60

35.40

6.65

2.10

1.39 1.79

1.29

1.05

1.96

11.37

13.02

5.71

2.65

Wind Speed ( kmh)

1 11 21 31 41 51

Joint Frequency Distribution

Port aux Basques, NL (1966-2004) June

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 10% intervals.Calms included at center.

3.92

2.18 0.94

1.07

4.13

37.82

7.15

2.18

1.54 1.66

1.41

1.09

2.11

12.27

13.13

5.27

2.12

Wind Speed ( kmh)

1 11 21 31 41 51

Joint Frequency Distribution

Port aux Basques, NL (1966-2004) July

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 10% intervals.Calms included at center.

3.99

1.32 0.66

0.80

4.41

39.06

7.16

2.42

1.67 2.25

1.70

1.35

2.74

13.16

11.82

3.88

1.62

Wind Speed ( kmh)

1 11 21 31 41 51

Joint Frequency Distribution

Port aux Basques, NL (1966-2004) August

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

5.70

2.30 1.17

1.34

5.89

24.99

5.39

2.77

2.78 4.15

2.83

2.18

3.21

14.63

12.21

5.84

2.62

Wind Speed ( kmh)

1 11 21 31 41 51

149

Joint Frequency Distribution

Port aux Basques, NL (1966-2004) September

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

5.04

4.13 1.80

1.99

5.66

17.76

4.54

2.76

2.76 4.05

2.78

2.22

3.03

13.44

15.08

8.82

4.14

Wind Speed ( kmh)

1 11 21 31 41 51

Joint Frequency Distribution

Port aux Basques, NL (1966-2004) October

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

3.12

4.62 2.34

2.12

5.87

15.22

3.60

2.47

2.65 3.97

2.97

2.71

3.96

15.44

13.69

10.02 5.21

Wind Speed ( kmh)

1 11 21 31 41 51

Joint Frequency Distribution

Port aux Basques, NL (1966-2004) November

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

1.98

4.65 2.86

2.87

5.75

13.56

3.42

2.28

2.59 3.98

3.01

2.50

3.95

14.64

14.30

11.93

5.73

Wind Speed ( kmh)

1 11 21 31 41 51

Joint Frequency Distribution

Port aux Basques, NL (1966-2004) December

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

1.75

5.24 3.29

3.35

6.48

9.95

3.29

2.22

2.29 3.49

2.83

2.53

4.47

13.92

15.64

13.14

6.12

Wind Speed ( kmh)

1 11 21 31 41 51

150

Joint Frequency Distribution

Stephenville A, NL (1967-2004) January

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

3.34

3.51 2.95

6.81

9.18

7.62

1.50

1.38

1.57 4.49

2.83

3.39

7.41

20.66

12.21

7.08

4.08

Wind Speed ( kmh)

1 11 21 31 41 51

Joint Frequency Distribution

Stephenville A, NL (1967-2004) February

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

5.64

4.18 2.19

8.09

10.56

7.50

1.45

1.11

1.26 4.36

2.50

4.00

8.39

17.94

10.54

6.29

4.01

Wind Speed ( kmh)

1 11 21 31 41 51

Joint Frequency Distribution

Stephenville A, NL (1967-2004) March

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

7.58

4.33 2.70

9.41

11.70

9.04

2.15

1.52

1.79 4.05

2.61

4.98 11.06

12.08

6.03

4.92

4.06

Wind Speed ( kmh)

1 11 21 31 41 51

Joint Frequency Distribution

Stephenville A, NL (1967-2004) April

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

6.86

4.50 3.80

8.18

13.07

12.49

3.58

2.17

1.92 3.29

2.40

5.52 11.07

8.51

4.35

4.44

3.84

Wind Speed ( kmh)

1 11 21 31 41 51

151

Joint Frequency Distribution

Stephenville A, NL (1967-2004) May

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

8.27

3.30 2.99

7.95

14.18

10.45

2.57

1.82

1.70 3.73

2.68

6.40 12.93

9.30

3.72

4.53

3.47

Wind Speed ( kmh)

1 11 21 31 41 51

Joint Frequency Distribution

Stephenville A, NL (1967-2004) June

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

11.10

2.57 3.00

6.86

9.52

8.13

2.28

1.44

1.94 4.29

3.03

6.65 16.27

11.96

3.49

4.37

3.10

Wind Speed ( kmh)

1 11 21 31 41 51

Joint Frequency Distribution

Stephenville A, NL (1967-2004) July

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

12.70

1.63 2.57

6.68

8.34

7.03

2.04

1.31

1.98 4.71

3.49

7.90 19.11

12.62

2.99

2.91

2.00

Wind Speed ( kmh)

1 11 21 31 41 51

Joint Frequency Distribution

Stephenville A, NL (1967-2004) August

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

9.88

2.39 3.19

7.21

8.65

6.49

1.33

1.22

1.90 5.36

4.15

7.52 18.15

12.14

3.87

3.89

2.66

Wind Speed ( kmh)

1 11 21 31 41 51

152

Joint Frequency Distribution

Stephenville A, NL (1967-2004) September

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

8.33

3.45 3.80 8.14

9.26

6.42

1.69

1.28

2.38

5.88

3.86

5.63 13.27

12.45

5.93

4.73

3.49

Wind Speed ( kmh)

1 11 21 31 41 51

Joint Frequency Distribution

Stephenville A, NL (1967-2004) October

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

6.25

4.59 3.87

7.82

8.86

7.15

1.82

1.44

2.28 5.44

3.57

4.81

10.24

14.90

7.05

5.62

4.30

Wind Speed ( kmh)

1 11 21 31 41 51

Joint Frequency Distribution

Stephenville A, NL (1967-2004) November

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

4.58

4.27 3.08

6.60

10.22

8.77

1.92

1.63

2.24 5.05

3.13

3.26

7.40

17.49

9.11

6.78

4.48

Wind Speed ( kmh)

1 11 21 31 41 51

Joint Frequency Distribution

Stephenville A, NL (1967-2004) December

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

4.42

4.33 3.63

6.57

9.85

9.21

1.80

1.31

1.57 4.40

2.38

2.30

6.96

18.69

11.54

6.59

4.44

Wind Speed ( kmh)

1 11 21 31 41 51

153

Joint Frequency Distribution

Sydney A, NS (1967-2004) January

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

2.51

5.54 2.37

2.07

1.84

3.12

1.77

2.51

3.09 7.19

7.51

9.90

11.04

19.48

9.45

5.81

4.80

Wind Speed ( kmh)

1 11 21 31 41 51

Joint Frequency Distribution

Sydney A, NS (1967-2004) February

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

2.57

9.11

2.86

2.91

2.57

3.50

2.02

2.27

2.71 6.91

7.08 9.07

8.89

17.53

8.84

5.82

5.35

Wind Speed ( kmh)

1 11 21 31 41 51

Joint Frequency Distribution

Sydney A, NS (1967-2004) March

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

2.69

11.66

3.82

3.62

3.10

3.55

2.44

2.70

3.60

8.51

8.12

8.56

7.76

12.19

6.52

5.33

5.84

Wind Speed ( kmh)

1 11 21 31 41 51

Joint Frequency Distribution

Sydney A, NS (1967-2004) April

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

2.76

12.08

5.85

4.27

3.80

4.86

3.82

3.83

3.85

8.22

7.91

7.53

6.36

9.44

4.92

4.85

5.66

Wind Speed ( kmh)

1 11 21 31 41 51

154

Joint Frequency Distribution

Sydney A, NS (1967-2004) May

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

3.04

11.91

4.71

4.13

3.63

4.22

3.12

3.39

4.27

10.90

11.12

9.68

6.36

7.11

3.65

3.70

5.06

Wind Speed ( kmh)

1 11 21 31 41 51

Joint Frequency Distribution

Sydney A, NS (1967-2004) June

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

4.03

8.86

3.49

2.96

2.49

3.32

2.69

3.27

4.03

11.50

14.14

14.09

8.06

7.75

3.37

2.76

3.19

Wind Speed ( kmh)

1 11 21 31 41 51

Joint Frequency Distribution

Sydney A, NS (1967-2004) July

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

4.11

4.86 2.83

2.32

1.69

2.78

2.18

2.51

3.80

13.21

17.26

16.71

9.71

8.26

3.32

2.34

2.11

Wind Speed ( kmh)

1 11 21 31 41 51

Joint Frequency Distribution

Sydney A, NS (1967-2004) August

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

4.64

5.45 2.80

2.60

2.07

3.26

2.11

2.24

3.32

10.22

14.43

16.76

11.34

9.53

3.67

2.88

2.68

Wind Speed ( kmh)

1 11 21 31 41 51

155

Joint Frequency Distribution

Sydney A, NS (1967-2004) September

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

4.50

6.99

2.79

2.70

1.95

3.43

2.13

2.33

3.25

9.57 11.24

13.80

10.52

10.75

5.79

4.33

3.93

Wind Speed ( kmh)

1 11 21 31 41 51

Joint Frequency Distribution

Sydney A, NS (1967-2004) October

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

3.52

6.90

3.14

2.90

1.75

2.90

2.37

2.67

3.54

8.68 10.25

12.81

10.24

12.09

5.99

5.30

4.96

Wind Speed ( kmh)

1 11 21 31 41 51

Joint Frequency Distribution

Sydney A, NS (1967-2004) November

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

3.36

6.54

3.16

2.76

2.33

3.14

2.06

2.49

3.49

9.05

8.47

9.37

10.04

14.37

8.08

5.97

5.31

Wind Speed ( kmh)

1 11 21 31 41 51

Joint Frequency Distribution

Sydney A, NS (1967-2004) December

N

S

W E

No observations were missing.Wind flow is FROM the directions shown.Rings drawn at 5% intervals.Calms included at center.

2.46

5.98 2.24

2.27

2.06

3.01

2.45

3.02

3.59 7.32

7.12

8.96

10.69

17.84

9.23

6.16

5.61

Wind Speed ( kmh)

1 11 21 31 41 51