reply from john prior, national climate information centre, met office, exeter

1
Weather – October 2007, Vol. 62, No. 10 291 Met Office Have other readers noticed that the Met Office have recently changed the historical context in which unusual months and sea- sons are viewed? They have drawn a line at 1914, and now pretend that all records taken before that simply does not exist. There may be good in-house reasons for this, although it is unlikely that they are relevant to the media or the general public. However, the decision is not (and has never to my knowledge been) explained. New records are now, by this method, much easier to obtain, but if events are labelled “the wettest since records began” or “unprecedented”, then the people we pay to keep our infrastructure running have a ready-made excuse for failure. “It’s never happened before, so how could we be expected to plan for it?” By so doing they have - probably uncon- sciously, though I could easily be persuaded otherwise – made my job as first port of call for many journalists on weather stories much more difficult than it ought to be. Philip Eden Whipsnade, Bedfordshire DOI: 10.1002/wea.131 Belief in records If the Met Office no longer believes records prior to 1914 are acceptable, thus making June 2007 the ‘wettest on record’, should we discount the global and regional warm- ing trend identified in the instrumental records since the mid-nineteenth century? Can we strike the ‘Little Ice Age’ from the textbooks? Was the meticulous work of observers in the eighteenth and nineteenth century (with often beautifully made, accu- rate equipment) a waste of time? Is my admiration for Professor Manley’s work on the Central England Temperature series, extending back to 1659, misplaced? Michael Dukes Chesham Buckinghamshire DOI: 10.1002/wea.132 Reply from John Prior, National Climate Information Centre, Met Office, Exeter These letters concern the areal series that may be used to place a UK weather event into context, so their main features will be described to help explain choices that are made for individual events and, in particu- lar, the significance of 1914. The series include the historic Central England Temperature (CET) and UK precipi- tation (HadUKP) series. The CET series has values representative of an area from the South Midlands to Lancashire, on a monthly basis from 1659 and daily from 1772. The HadUKP series represent England and Wales, Scotland and Northern Ireland, together with 8 sub-regions, with various starting dates e.g. 1766 for England and Wales (combined) and 1931 for Scotland and Northern Ireland. These historic series have been calculated using a limited set of long period stations, with the intention of preserving homogeneity. A further type of series has been devel- oped by the Met Office more recently, based upon all station data that are avail- able electronically. For a given element, these have been interpolated to 5km grid points and then areal values produced by averaging the relevant gridded values. In this way it has been possible to produce areal series for the UK, constituent coun- tries, climate districts and counties. In gen- eral, these series start in 1961 but, as a result of a programme of digitising data published in the Monthly Weather Report (MWR), the sunshine series extends back to 1929 and the temperature and precipitation ones to 1914. Why these particular years? A relative- ly dense, even station network is needed for creating gridded data sets, and 1929 is the earliest year with a suitable sunshine station network. 1914 is the year when the format of the MWR changed to one that allowed efficient digitization for the period to 1960. The feasibility of further extending the grid- ded series will be investigated, e.g. to 1900 or even to 1884 when the MWR was first published. In addition, a homogeneous merger of the gridded and historic (HadUKP) precipitation series is being pur- sued for some of the areas, to enable only the gridded type to be maintained from now on. The intention is to publish a peer- reviewed paper about this work. Whilst being shorter than most of the his- toric series, the gridded series have the advantage of being available for many more sub-divisions of the UK - from individual counties upwards. So when the heat of a month such as July 2006 needs to be placed into context, that may be done for all parts of the UK not just central England and when flooding occurs in Yorkshire as it did in June 2007, a ranking in the Yorkshire precipita- tion series can be referred to. The gridded series have been quoted regularly in the annual and seasonal summaries published in Weather, beginning with those for 2005 and winter 2005/06 (Weather, 61:4), and are used in parts of Weather Log. The UK temperature and precipitation series from 1914 may be used to satisfy media demand for a greater range of infor- mation, particularly about local areas. These series also enable the weather to be placed in a ‘within living memory’ context. Howev- er, when a weather event warrants it, the longer series are also used. Recent examples include April 2007 being the warmest April in the 349-year CET series (Met Office News Release, 2 May 2007) and May to July 2007 being the wettest such period over England and Wales in the 241-year HadUKP series (Met Office News Release, 26 July 2007). Further information about the two types of series is available at http://www. metoffice.gov.uk/climate/uk/about/ methods.html John Prior Exeter DOI:10.1002/wea.161 The Met Office News Release of 5 July 2007 entitled ‘Record breaking June rainfall figures’ stated ‘the month of June has been the wettest since records began in 1914’. This date refers to the start date of the Met Office areal rainfall series rather than the station-based England and Wales Rainfall series that starts in 1766. The Met Office Seasonal Weather Summary for summer 2007 which appears on page 276 of this issue refers to both rainfall series. (Ed.) Letters

Upload: john-prior

Post on 06-Jul-2016

213 views

Category:

Documents


0 download

TRANSCRIPT

Weather – October 2007, Vol. 62, No. 10

291

Met Office

Have other readers noticed that the MetOffice have recently changed the historicalcontext in which unusual months and sea-sons are viewed?

They have drawn a line at 1914, and nowpretend that all records taken before thatsimply does not exist. There may be goodin-house reasons for this, although it isunlikely that they are relevant to the mediaor the general public. However, the decisionis not (and has never to my knowledgebeen) explained.

New records are now, by this method,much easier to obtain, but if events arelabelled “the wettest since records began”or “unprecedented”, then the people wepay to keep our infrastructure running havea ready-made excuse for failure. “It’s neverhappened before, so how could we beexpected to plan for it?”

By so doing they have - probably uncon-sciously, though I could easily be persuadedotherwise – made my job as first port of callfor many journalists on weather storiesmuch more difficult than it ought to be.

Philip EdenWhipsnade,

BedfordshireDOI: 10.1002/wea.131

Belief in records

If the Met Office no longer believes recordsprior to 1914 are acceptable, thus makingJune 2007 the ‘wettest on record’, shouldwe discount the global and regional warm-ing trend identified in the instrumentalrecords since the mid-nineteenth century?Can we strike the ‘Little Ice Age’ from thetextbooks? Was the meticulous work ofobservers in the eighteenth and nineteenthcentury (with often beautifully made, accu-rate equipment) a waste of time? Is myadmiration for Professor Manley’s work onthe Central England Temperature series,extending back to 1659, misplaced?

Michael DukesChesham

Buckinghamshire

DOI: 10.1002/wea.132

Reply from John Prior, NationalClimate Information Centre, MetOffice, Exeter These letters concern the areal series thatmay be used to place a UK weather eventinto context, so their main features will bedescribed to help explain choices that aremade for individual events and, in particu-lar, the significance of 1914.

The series include the historic CentralEngland Temperature (CET) and UK precipi-tation (HadUKP) series. The CET series hasvalues representative of an area from theSouth Midlands to Lancashire, on a monthlybasis from 1659 and daily from 1772. TheHadUKP series represent England andWales, Scotland and Northern Ireland,together with 8 sub-regions, with variousstarting dates e.g. 1766 for England andWales (combined) and 1931 for Scotlandand Northern Ireland. These historic serieshave been calculated using a limited set oflong period stations, with the intention ofpreserving homogeneity.

A further type of series has been devel-oped by the Met Office more recently,based upon all station data that are avail-able electronically. For a given element,these have been interpolated to 5km gridpoints and then areal values produced byaveraging the relevant gridded values. Inthis way it has been possible to produceareal series for the UK, constituent coun-tries, climate districts and counties. In gen-eral, these series start in 1961 but, as a resultof a programme of digitising data publishedin the Monthly Weather Report (MWR), thesunshine series extends back to 1929 andthe temperature and precipitation ones to1914. Why these particular years? A relative-ly dense, even station network is needed forcreating gridded data sets, and 1929 is theearliest year with a suitable sunshine stationnetwork. 1914 is the year when the formatof the MWR changed to one that allowedefficient digitization for the period to 1960.The feasibility of further extending the grid-ded series will be investigated, e.g. to 1900or even to 1884 when the MWR was firstpublished. In addition, a homogeneousmerger of the gridded and historic(HadUKP) precipitation series is being pur-sued for some of the areas, to enable only

the gridded type to be maintained fromnow on. The intention is to publish a peer-reviewed paper about this work.

Whilst being shorter than most of the his-toric series, the gridded series have theadvantage of being available for many moresub-divisions of the UK - from individualcounties upwards. So when the heat of amonth such as July 2006 needs to be placedinto context, that may be done for all partsof the UK not just central England and whenflooding occurs in Yorkshire as it did in June2007, a ranking in the Yorkshire precipita-tion series can be referred to. The griddedseries have been quoted regularly in theannual and seasonal summaries publishedin Weather, beginning with those for 2005and winter 2005/06 (Weather, 61:4), and areused in parts of Weather Log.

The UK temperature and precipitationseries from 1914 may be used to satisfymedia demand for a greater range of infor-mation, particularly about local areas. Theseseries also enable the weather to be placedin a ‘within living memory’ context. Howev-er, when a weather event warrants it, thelonger series are also used. Recent examplesinclude April 2007 being the warmest Aprilin the 349-year CET series (Met Office NewsRelease, 2 May 2007) and May to July 2007being the wettest such period over Englandand Wales in the 241-year HadUKP series(Met Office News Release, 26 July 2007).

Further information about the two types of series is available at http://www.metoff ice .gov.uk/c l imate/uk/about/methods.html

John PriorExeter

DOI:10.1002/wea.161

The Met Office News Release of 5 July 2007entitled ‘Record breaking June rainfallfigures’ stated ‘the month of June has beenthe wettest since records began in 1914’.This date refers to the start date of the MetOffice areal rainfall series rather than thestation-based England and Wales Rainfallseries that starts in 1766. The Met OfficeSeasonal Weather Summary for summer2007 which appears on page 276 of thisissue refers to both rainfall series. (Ed.)

Letters