Temperature and Humidity Effects on Particulate Matter

Concentrations in a Sub-Tropical Climate During Winter

German Hernandez1, Terri-Ann Berry

1, Shannon L Wallis

1 & David Poyner

1

Abstract. High concentrations of particulate matter (PM) in the atmosphere have been associated with

degradation of human health. Auckland is the most populous city in New Zealand, with a humid subtropical

climate (characterised by hot and humid summers and mild winters) and is one of the most remote cities in

the world. Analysis of the weather conditions and PM10 concentrations in Auckland can provide an

understanding of how weather conditions impact PM10 concentrations in isolated systems. This study looks

at the impacts of Temperature and Relative Humidity (RH) on localised PM10 concentrations in an urban

area of Auckland over the winter period.

This study has shown that the ratios of PM1:PM10 and PM2.5:PM10 were 64% and 87% over the study

period. Temperature was observed to have a negative correlation with PM10 over a diurnal timescale. RH

generally showed a positive correlation with PM10 up to a threshold value of 75% RH, beyond which the

correlation ceased. RH affects the natural deposition process of PM, whereby moisture particles adhere to

PM, accumulating atmospheric PM concentration. With increasing humidity, moisture particles eventually

grow in size to a point where ‘dry deposition’ occurs, reducing PM10 concentrations in the atmosphere.

There were a number of occasions where PM10 remained low even while RH increased and further

investigation showed these events coincided with periods of rainfall.

Keywords: particulate matter, weather conditions, pollution

1. Introduction

Air pollution measured as particulate matter (PM) has been shown to be detrimental to human health [1],

[2]-[4] and lead to increased mortality rates [5], [6]. World Health Organisation (WHO) recommends a long-

term guideline limit of 20 µg/m3 PM10 (annual average) to provide a minimum level of protection against

long-term health risk [7].

Although some studies suggest that PM2.5 is the more harmful fraction of PM10 [8], [9] in New Zealand,

PM10 is the major air pollutant monitored to provide an indicator of general air quality conditions. The NZ

National Environmental Standards (NES) for Air Quality have adopted the WHO guideline limit for PM10

(with a daily limit for PM10 of 50 µg/m3) [10]. In 2013, a review of annual average PM10 concentrations

across New Zealand indicated that there were eight areas which exceeded the NES limit [11]. Of these, three

regions were in excess by 21-35% and 5 regions by 1-10%.

As the largest city in NZ, Auckland has some areas that occasionally exceed the annual average PM10

limit (4 out of 8 years between 2006 and 2013 recorded PM10 levels between 21 and 30 g/m3, [12].

Auckland is predicted to reach a population of 2 million by the early 2030s [13] and accordingly there is no

certainty that the PM10 limit will not be exceeded again in the future.

Auckland is located in the North Island of New Zealand and has a population of around 1.5 million, 32%

of the country’s population. The metropolitan area has a population density of 2,700 people per square

kilometre. According to the 2013 Census there were around 470,000 households in Auckland [14]. The

number of cars on Auckland’s roads was 1.02 million [15] Currently, 85% of trips in Auckland are made by

private car, and around 15,000 extra cars join Auckland’s roads every year [16]. Auckland has a subtropical

Corresponding author. Tel.: +64022 137 1876

E-mail address: hhernandez@unitec.ac.nz

International Proceedings of Chemical, Biological and Environmental Engineering, V0l. 102 (2017)

DOI: 10.7763/IPCBEE. 2017. V102. 8

41

climate with warm, humid summers and mild, damp winters. Table 1 provides a summary of key weather

data for Auckland [17].

Table 1: Average Weather Statistics for Auckland

Average summer daily temperature 23oC Mean annual rainfall 1,240 mm

Average winter daily temperature 14oC Mean annual wind speed 3.9 m/s

Due to Auckland’s isolated position in the South Pacific, air arriving at the coastline is relatively pure

and unpolluted [18]. However, numerous natural and anthropogenic inputs, mainly from mainland and

coastal areas, can deteriorate air quality. The primary sources of PM10 in Auckland during winter are as

shown in Table 2. PM10 levels are nearly four times higher in winter than in summer [19].

Table 2: Primary Sources of PM10 in Auckland

Source Examples % Contribution during Winter

Domestic Wood fires during winter 72%

Industry Metal finishing, mining, construction 7%

Transport Domestic and public 21%

Various studies have shown that PM10 concentration is not simply a product of emission source

concentration, but is also influenced by diffusion conditions. Diffusion of particulate matter is in turn

affected by both geographical and meteorological conditions. Meteorological conditions that can affect

PM10 include wind, temperature, precipitation and relative humidity [20]-[22].

In this article, the influence of Temperature and Relative Humidity on PM10 concentrations in

Auckland has been studied. Variations in PM10 concentrations were analysed during the winter months,

July–September 2016. This data was used to determine how PM10 concentrations in Auckland were affected

by the meteorological conditions.

2. Materials and Methods

Auckland is located in an isthmus in the northern part of New Zealand, as can be seen in the map in Fig.

1. The city has a humid, subtropical climate with warm, humid summers and mild winters.

Auckland is surrounded by around 3,100 kilometres of coastline [23] with the Tasman Sea to the west

and the Pacific Ocean to the east. The only developed neighbouring country to Auckland is Australia, more

than 1,700 kilometres away. PM concentrations occurring in Auckland are therefore more likely to originate

from within the country than from overseas sources.

Fig. 1: Map showing Auckland and its proximity to Australia

42

Unitec has over 16,000 students. The main campus is located on a 55 hectare site within Mount Albert,

an inner city suburb centred on a volcanic peak. The campus is surrounded by residential housing and is

located approximately six kilometres southwest of the Central Business District (CBD). The southern

coastline of the Waitemata Harbour is situated approximately 0.7 km to the west of the Unitec campus. The

campus is bordered by Oakley Creek along its western side for 1.3 km. The creek separates the campus site

from Great North Road which is a major traffic thoroughfare to the western suburbs of Auckland. The

northern boundary of the campus is bordered by the Northwestern Motorway over approximately 200 metres.

The Northwestern Motorway is the major western route and secondary northern route out of Auckland.

Data was collected over the eight week period during July to September 2016. Analysis was carried out

at 5 minute intervals for the testing period. Particulate matter analysis was performed by two dust profilers

(Aeroqual AQM PP8) to provide continuous and simultaneous measurement of PM10, PM2.5, PM1, and

TSP (Total Suspended Particulates). Each profiler comprised an optical particle counter which converts

counts to mass fraction via a proprietary algorithm. The sensor head range is PM1 0-200 µg/m3; PM2.5 0-

2000 µg/m3; PM10 0-5000 µg/m3; TSP 0-5000 µg/m3, with accuracy of < ±5 µg/m3 + 15% of reading. The

profilers were located at Unitec, at 3.3 metres height above the ground. The distance to the nearest building

was approximately 15 metres. This location was chosen both for minimal influence from nearby buildings

and to avoid tampering or vandalism. The profiler comprises an optical particle counter that converts counts

to a mass fraction via a proprietary algorithm stored in the system firmware (Aeroqual, Auckland, NZ).

Weather station data was obtained from an on-site weather station, mounted 10m above ground level.

Details of the weather data monitoring equipment is summarised in Table 3.

Table 3: Weather Station Equipment Details

Parameter Monitoring equipment Range Accuracy

Wind Vaisala WXT530 0 to 60 m/s < ±3% at 10m/s

Relative Humidity Vaisala WXT530 0 to 100% RH < ±3% RH at 90% RH

Temperature Vaisala WXT530 -52 to +60°C < ±0.3°C at 20°C

Rainfall OTA OSK15180T tipping bucket rain

gauge2

N/A <2% at rainfall intensity of

100mm/hr

3. Results & Discussion

Concentrations of particulate matter in the atmosphere (PM1, PM2.5 and PM10) were monitored on site

at the Unitec campus over an eight week study period, together with corresponding atmospheric conditions.

The resulting weekly averages were calculated over the study period. Table 4 compares the maximum,

average and minimum values with the highest and lowest weekly averages.

The New Zealand National limit for annual average PM10 values, which is based on the WHO guideline

limit, is 20 µg/m3. From Table 4 it can be seen that average weekly PM10 values remained below this limit

throughout the study period. This could be attributed to New Zealand’s generally high level of air quality or

Unitec’s location within a relatively suburban residential area, away from highly polluting sources, such as

the city centre.

The daily average PM10 concentrations were also calculated and ranged between 1.91 µg/m3 to 10.72

µg/m3 over the study period, which is less than the National Environmental Standards of 50 µg/m3.The

relative proportions (based on weekly averages) of particulate matter (PM10, PM2.5 and PM1) are shown in

Fig. 2. The PM2.5:PM10 ratio was 87% while the PM1:PM10 ratio was less, at 64%. Previous findings

identified ratios of 72% and 61% respectively [24] and 64% for PM2.5:PM10 [25], which may indicate

slightly elevated relative levels of PM2.5, possibly due to seasonality and the use of solid fuel burners in

Auckland.

The weekly average concentrations of PM1, PM2.5 and PM10 were all observed to follow similar trends

across the study period. Spearman’s Rank Analysis was used to determine the degree of correlation between

43

PM10 and other particle sizes. Comparing PM1 and PM2.5 values with PM10 values across the study period,

the Spearman’s Rank Coefficient was 0.9686 and 0.9925 for PM1 and PM2.5 respectively. This is consistent

with the findings of Li [24].

As the relative proportions of particulate matter fractions remained fairly constant during this study and

also because PM10 is the main air pollutant monitored in New Zealand, this study will focus on the analysis

of PM10 with respect to meteorological impacts.

Table 4: Summary of air quality measurements and ambient meteorological conditions on campus

PM10

(µG/M3)

PM2.5

(µG/M3)

PM1

(µG/M3)

TEMP

(°C)

RH

(%)

Max Value 74.74 73.16 62.97 19.6 92.7

Highest Weekly Average 8.73 7.73 5.62 13.76 77.11

Average 6.01 5.22 3.84 12.1 74.48

Lowest Weekly Average 3.75 3.26 2.51 9.49 66.97

Min Value 0.48 0.31 0.18 3.7 40.4

Fig. 2: Relative proportions of various size fractions of particulate matter

PM10 concentrations were further analysed to establish a weekly pattern, as shown in Fig. 3, to help

understand reasons for variability. As expected, higher concentrations were observed during the week, when

most activity occurs on campus. The levels reached a maximum on Fridays and were lowest on Saturdays.

This could be explained by the fact that many Aucklanders like to “get out of the city” for the weekend,

leaving Friday evening and returning Sunday.

Fig. 3: Daily trends in average [PM10]

44

Fig. 4 shows the average concentration of PM10, across five separate timeframes (e.g. early morning

from 0.00 – 5.55 hr) for each of the days of the week. The diurnal profiles were fairly consistent, with one

exception observed on Tuesdays. The average on Tuesdays was remarkably high during the late evening

(19:00 – 23:55). The authors cannot find any likely explanation for this observation. However, there is a

laundry service on campus (Taylors) which may have a particular operation on this day.

The ratio of PM2.5:PM10 was also analysed over the same time periods, averaged for each of the days of

the week. There was a consistent trend observed throughout the week, with the ratio at its highest during the

early morning and late evening (average 89% and 91% respectively) and lowest around the middle of the day

(average 83%). This could be due to the increased traffic volumes that tend to occur during these periods.

Fig. 4: Average PM10 values (g/m3) over five time periods (during 24hrs) for each day

Fig. 5 also shows that there is a morning peak in PM10 concentrations between 7am and 9am, coinciding

with the peak morning traffic period. Between 5pm and 7pm, there is a rapid increase in PM10

corresponding to the evening peak traffic period. An even sharper increase in PM10 concentrations is

observed from 7pm, with concentrations remaining high until 11pm. This can be explained by increased

wood burning during the winter evenings.

Fig. 5: Daily Variation in PM10 Concentration (typical week day)

For comparison, Figure 6 shows daily variation in PM10 concentration across a typical Saturday during

the study period. As expected, the peak morning increase observed during the weekday diurnal profile does

45

not appear on the weekend graph. However, an evening peak in PM10 is observed, which again may be

explained by an increase in wood burning.

3.1. Temperature

Temperature, together with PM10 concentrations, was studied over a 24 hour period to determine the

degree of influence temperature has on PM10 levels throughout a typical day.

Fig. 6: Daily Variation in PM10 Concentration (typical Saturday)

The diurnal profiles in Fig. 7 show there is a negative correlation between the two parameters.

Temperatures reach a maximum during the daytime while PM10 concentrations tend to be at their lowest.

Conversely, while temperatures reach a minimum overnight, PM10 levels are at their highest. Spearman’s

Rank analysis was used to verify the degree of correlation. The Spearman’s Coefficient was calculated to be

-0.45, indicting a moderately negative correlation. This finding is consistent with those of Barmpadimos [20]

The effect of increased daytime temperatures on PM10 concentrations may be explained by the process

of thermally-induced convection. As the ground heats up during the day, gusts and winds increase, leading to

increased diffusion of particulate matter [26]

The negative correlation observed at night time may be due to a number of factors. At night, decreasing

temperatures create a temperature inversion which acts as a cap, inhibiting diffusion of particulate matter.

Condensation of volatile compounds and increased wood burning are other possible causes for increased

PM10 concentrations at night. According to Auckland Council (2015), burning wood for home heating is the

primary contribution to pollution concentrations during winter.

Fig. 7: Daily Variations of PM10 concentration with ambient external temperature

3.2. Relative Humidity

46

A positive correlation was observed between the daily variations in Relative Humidity (RH) and PM10

concentrations. As shown in Fig. 8, higher values were observed during the night time and early morning,

with lower values occurring during the afternoons. Spearman’s rank coefficient for the period shown is 0.6,

indicating a strong, positive correlation.

Although the diurnal profile shows a positive correlation between RH and PM10, a lag between when

PM10 concentrations fall and when RH falls was observed. Specifically, while both PM10 and RH increase

up to a point, PM10 suddenly drops dramatically for a period while RH continues to climb and then falls

around 9 to 12 hours later, before repeating the pattern again. This rapid drop in PM10 tends to coincide with

when RH reaches a value of around 75%.

Barmpadimos [20] made a similar observation, stating that while PM2.5 concentrations increased while

humidity was low, once humidity reaches a high enough value, PM2.5 concentrations decreased. This

decrease is attributed to the particles accumulating mass due to moisture which ultimately leads to dry

deposition of the particles to the ground.

It was noted that there were some periods where this correlation did not occur at all. While RH continued

on a diurnal pattern, there were periods where the level of PM10 concentrations appeared to be suppressed.

On further investigation, it was found that these events occurred during periods of significant rainfall. This

may be explained by the process of wet deposition, whereby particulate matter is removed from the

atmosphere by inclusion or solution in precipitation, causing PM10 levels to decrease, while RH increased.

Fig. 8: PM10 variations with RH over a four-day timeframe

4. Conclusions

Meteorological conditions and PM levels in Auckland were monitored over an eight week period in

winter. The results were analysed to establish relationships between RH and Temperature with PM.

There was a strong relationship between the concentrations of PM for different particle sizes.

Spearman’s coefficient for the correlation between PM1 and PM10 was 0.9686, and 0.9925 for the

correlation between PM2.5 and PM10. The average weekly PM10 values over the study period remained

below the NZ national limit of 20 µg/m3. The daily average values ranged between 1.91 µg/m3 to 10.72

µg/m3, which is less than the National Environmental Standards of 50 µg/m3.

Temperature was observed to have a negative correlation with PM10 over a diurnal timescale. RH

generally showed a positive correlation with PM10 up to a threshold value of 75% RH, beyond which the

correlation ceased. There were a number of occasions where PM10 remained low even while RH increased

and further investigation showed these events coincided with periods of rainfall.

It was realised during the course of this study that it can be difficult to analyse meteorological variables

in isolation. Future studies will consider the additional impacts of wind speed, wind direction and rainfall on

PM10 concentrations. As these variables form part of a complex system the combined effects of these

variables will also be considered. Future studies will also consider analysis over a longer time period.

47

5. Acknowledgments

The authors wish to thank Building Research Association of New Zealand (BRANZ) for providing

funding for this project. Also thanks to Andrew Schunke (Chester Consultants, Auckland), Gregor Steinhorn

(Unitec, Research Pathway) and Surya Narain (Unitec, Engineering Pathway) for their invaluable assistance.

6. References

[1] Cohen, H. Anderson, B. Ostra, K. Pandey, M. Krzyznowski, N. Künzli, K.Gutschmidt,, A. Pope,I. Romieu,J.

Samet, K. Smith. The global burden of disease due to outdoor air pollution. J. Toxicol. Environ. Health. 2005, 68:

1-7 [2] K. Donaldson, V. Stone, A. Seaton, W. Mac Nee. Ambient particle inhalation and the cardiovascular system:

potential mechanisms. Environ. Health Persp. 2001, 109: 523-527.

[3] C. Pope, D. Bates, M. Raizenne. Health effects of particulate air pollution: time for reassessment? Environ. Health

Persp. 1995, 103: 472–480.

[4] R. Puett, J. Hart, J. Yanosky, D. Spiegelman, M. Wang, J. Fisher, B. Hong, F. Laden. Particulate Matter Air

Pollution Exposure, Distance to Road, and Incident Lung Cancer in the Nurses’ Health Study Cohort. Environ.

Health Persp. 2014, 122: 926-932.

[5] D. Dockery, C, Pope, X. Xu, J. Spengler, J. Ware, M. Fay, B. Ferris, F. Speizer. An association between air

pollution and mortality in six U.S. cities. N. Engl. J. Med. 1993, 329: 1753-1759.

[6] S. Hales, T. Blakely, A. Woodward. Air Pollution and mortality in New Zealand: cohort study. J Epidemiol

Community Health. 2010 66: 468-473.

[7] World Health Organization. WHO Air quality guidelines for particulate matter, ozone, nitrogen dioxide and sulfur

dioxide. Global update 2005. 2006 pp8,

[8] B. Ostro, J. Hu, D. Goldberg, P. Reynolds, A. Hertz, A. Bernstein, M. Kleeman. Associations of mortality with

long-term exposures to fine and ultrafine particles, species and sources: Results from the California teachers study

cohort. Environ. Health Persp. 2015, 123: 549-556.

[9] N. Hodas, Q. Meng, M. Lunden, D.Rich, H. Özkaynak, L. Baxter, B. Turpin. Variability in the fraction of ambient

fine particulate matter found indoors and observed heterogeneity in health effect estimates. J. Expo. Sci. Environ.

Epidemiol. 2012, 22: 448-454.

[10] Ministry for the Environment (New Zealand). Ambient Air Quality Guidelines, 2002 Update. 2012. Available:

http://www.mfe.govt.nz/sites/default/files/ambient-guide-may02.pdf

[11] Environmental Health Indicators New Zealand. Particulate matter. 2017. Available:

http://www.ehinz.ac.nz/indicators/air-quality/particulate-matter/

[12] Statistics New Zealand. Annual average PM10 concentrations in towns and cities. 2015a. Available:

http://www.stats.govt.nz/browse_for_stats/environment/environmentalreporting-series/environmental-

indicators/Home/Air/annual-pm10-concentrations-townscities.aspx

[13] Statistics New Zealand. Population growth in all regions, at least in the short term. 2015b. Available:

http://www.stats.govt.nz/browse_for_stats/population/estimates_and_projections/SubnationalPopulationProjection

s_MR2013base.aspx

[14] Auckland Council. Auckland Profile – Initial results from the 2013 Census. Social and Economic Research Team,

Investigations and Monitoring Unit, Auckland Council, Auckland, 2014.

[15] Ministry of Transport. Transport volume: Fleet information. 2017. Available:

http://www.transport.govt.nz/ourwork/tmif/transport-volume/tv004/

[16] Auckland Council. Auckland’s Transport. 2017b. Available: http://theplan.theaucklandplan.govt.nz/aucklands-

transport/

[17] National Institute of Water and Atmospheric Research. The climate and weather of Auckland. 2014. Avalaible:

https://www.niwa.co.nz/static/Auckland%20ClimateWEB.pdf

[18] Auckland Council. Managing Auckland’s air quality. 2017c. Available:

http://www.aucklandcouncil.govt.nz/EN/environmentwaste/pollution/Pages/airquality.aspx

[19] Auckland Council. The health of Auckland’s natural environment 2015. 39-47.

[20] I. Barmpadimos, C. Hueglin, J. Keller, S. Henee, A, Précȏt. Influence of meteorology on PM10 trends and

variability in Switzerland from 1991 to 2008. Atmo. Chem. Phys. 2011, 11: 1813-1835.

[21] D. Giri, V. Krishna, P.R. Adhikary. The influence of meteorological conditions on PM10 concentrations in

Kathmandu Valley. Int. J. Environ. Res. 2008, 2 (1): 49-60.

[22] P.Hien, V. Bac, H, Tham, D. Nhan, L. Vinh. Influence of meteorological conditions on PM2.5 and PM2.5-10

concentrations during the monsoon season in Hanoi, Vietnam. Atmos. Environ. 2002, 36: 3473-3484.

[23] Auckland Council. Auckland’s Coastline. 2017a. Available:

http://www.aucklandcouncil.govt.nz/EN/environmentwaste/coastalmarine/Pages/aucklandscoastline.aspx

[24] Y. Li, Q. Chen, H. Zhao, L. Wang, R. Tao. Variations in PM10, PM2.5, and PM1.0 in an urban area of the

Sichuan Basin and their relation to meteorological factors. Atmosphere. 2015, 6: 150-163.

[25] S. Munir, T. Habeebullah, A. Mohammed, E. Morsy, M. Rehan, K. Ali. Analysing PM2.5 and its association with

PM10 and meteorology in the arid climate of Makkah, Saudi Arabia. Aerosol Air Qual. Res. 2017, 453-464.

48

[26] I. Pepper, C. Gerba, M. Brusseau. Environmental and pollution science, second ed. Academic Press, Canada, 2006.

49