_______________________________________________________________________________ EMULATE: land station daily MSLP series _______________________________________________________________________________ 86 Land station daily MSLP series were included in the final gridded MSLP product, EMSLP3. Two MSLP series for each of the 86 stations are provided here. These correspond to the 'uncorrected' and 'corrected' series. The corrections will be described in detail below. Firstly, the format of the files is described. ______________________________________________________________________________ 1. Format Each file is in ascii format. There are two lines of header information. The first line is the station name (including the directory). The second line contains some important station details: station_number, first_year, last_year, latitude, longitude, station height eg. For La Coruna 8001 1865 2002 4316 -850 58 The series then follows with year, month, day, pressure Notes: * the pressure is at mean sea level and is given in hPa. * missing MSLP values are given as -999.90 * if there is no station number, it is given 99999 * station height is given in metres * if no details of station height were available, we used the height given in the WMO records * if a change of height was known, we took an average of the heights. * latitudes are in degrees (not degree and minutes) x 100. For example: 45 degrees N, 30 minutes = 45.5 x 100 => 4550 43 degrees N, 15 minutes = 45.25 x 100 => 4525 * longitudes are in degrees (not degrees and minutes) x 100. West is denoted by negative values. For example, 30 degrees E, 12 minutes = 30.2 x 100 => 3020 10 degrees W, 55 minutes = -10.92 x 100 => -1092 _______________________________________________________________________________ 2. 'Uncorrected' Series These files are denoted by the file ending : .ts.asc For example for Malta the filename is: emulate_malta.ts.asc Although they are classed here as 'uncorrected', many of these stations have undergone considerable quality control (eg. WASA, IMPROVE stations, Durham etc). We call them 'uncorrected', because we (at the Hadley Centre) have applied minimal corrections - they are basically in the form that we received the data in. Minimal corrections refer to correcting typos, calculating the daily average, filling in missing data points in a couple of stations (using a different source) and reducing Hohenpeissenberg, Jena, Milan and Athens to mean sea level. As other groups have spend considerable time working on these series (a big thanks to David Lister, Manola Brunet-India, Anders Moberg etc), it is appropriate to give the main point of contact for each of these 'uncorrected' series. The name given below does not always correspond to the original source of the data, however they can probably point you in the right direction! aberdeen: WASA & R. Allan london : Rob Allan alexandria: R. Allan lugansk: D. Lister algiers: R. Allan lund: WASA ammasalik: WASA lyon: R. Allan angra: R. Allan madrid: M. Brunet-India archangel: D. Lister malta: D. Lister armagh: R. Allan milan: IMPROVE astrakhan: D. Lister montreal: V. Slonosky athens: R. Allan & D. Founda moscow: D. Lister baghdad: R. Allan nikolaef: D. Lister barcelona: M. Brunet-India nordby: WASA & A. Moberg beirut: R. Allan oksoyfyr: WASA & A. Moberg bergen: WASA & A. Moberg orenburg: D. Lister bermuda: D. Lister padua: IMPROVE biarritz: R. Allan palermo: M. Barriendos biskra: R. Allan paris: M. Barriendos bodo: WASA & O. Nordli & D. Lister plymouth: R. Allan brest: R. Allan potsdam: WASA cadiz: M. Brunet-India & IMPROVE prague: D. Lister corfu: D. Lister providence: D. Lister debilt: T. Brandsma reykjavik: P. Jones diarbekir: R. Allan riga: D. Lister durham: D. Lister rochefort: R. Allan fao: R. Allan rome: R. Allan funchal: R. Allan scutari: D. Lister galway: R. Allan (K. Hickey) sevastopol: D. Lister gibraltar: M. Rodwell & D. Wheeler sibiu: D. Lister godthaab: R. Allan stjohns: V. Slonosky goteborg: WASA & A. Moberg stockholm: WASA & A. Moberg halifax: D. Lister & V.Slonosky stornoway: R. Allan hammerodde: WASA stpetersburg: D. Lister haparanda: WASA & A. Moberg stykkisholmur: WASA harnosand: WASA tenerife: M. Brunet-India helsinki: WASA tiflis: D. Lister hohenpeissenberg: J. Jacobeit torshavn: WASA & A. Moberg jena: T. Ansell & F. Gerstengarbe toulon: R. Allan kazan: D. Lister uppsala: WASA & A. Moberg kem: D. Lister valentia: WASA & R. Allan kiev: D. Lister vardo: O. Nordli & D. Lister kostroma: D. Lister vestervig: WASA lacoruna: M. Brunet-India visby: WASA & A. Moberg lesina: R. Allan wilna: D. Lister lisbon: R. Allan zagreb: L. Srnec _______________________________________________________________________________ 3. 'Corrected' These files are denoted by the file ending : .ts.diurnal.adj.asc For example for Malta the filename is : emulate_malta.ts.diurnal.adj.asc Firstly we apply a correction for the diurnal and semi diurnal oscillation. We use the phase and amplitude fields (seasonal) calculated by Dai and Wang (1999) to estimate the diurnal & semi diurnal cycle at each station. Observation hours for each day of a station series have been produced and these are then used to determine the appropriate adjustment, given the cycle. Due to a lack of sufficient sub-daily data, we were unable to calculate the diurnal and semidiurnal cycle at each station directly. The Dai and Wang fields are on a 5x5 degree latitude longitude grid and hence we interpolate the nearest grid point to our station. It is rather a crude adjustment, however corrections are generally small and, as stated above, it was not possible to estimate diurnal waves directly for each station in this project. Secondly, we adjust the series such that the monthly mean is equal to a reference monthly mean (either the corresponding ADVICE station series, that interpolated from the nearest ADVICE grid point, or from the nearest HadSLP grid point). Specifically, the monthly means are calculated from the daily series and compared to the reference ADVICE/HadSLP monthly value. The difference in monthly means is then used to adjust the daily SLP values. There was some concern about jumps in adjustments at the end of each month, so a slight smoothing was applied to the monthly adjustments. Preference was given to the ADVICE station series where possible, however in some cases problems were found with the ADVICE station series. In these instances, the ADVICE gridded value was used. For the Canadian stations and those in the far east of the EMULATE region a final adjustment was required so that their daily average represented the same 24 hour period as the other series. The Canadian stations are 5 hours behind, the Russian 4 hours ahead. As only daily averages were available, we were forced to interpolate between the preceding (following) day and the actual day for the Canadian (far Russian) stations. These are denoted as : ts.diurnal.adj.gmt.ts Finally after all quality control checks were performed, a number of remaining gross outliers were identified. These were either corrected or replaced with missing values. This effected 4 stations and these are denoted: ts.diurnal.adj_g_error_corr.ts.asc eg. emulate_gibraltar.ts.diurnal.adj_g_error_corr.ts.asc ________________________________________________________________________________ 4. Remaining issues We do still have come concerns about the quality of a number of series. To properly quality control 86 stations would require more than time allocated to the whole of the EMULATE project. Because the data would undergo further quality control in the gridding procedure, we decided to proceed with the stations series as they stood. For research being conducted with individual station series, we would recommend however that some caution to be exercised. Any feedback would be extremely useful! Tara Ansell tara.ansell@metoffice.com ________________________________________________________________________________ 5. References Dai, A. and Wang, J. (1999) 'Diurnal and Semidiurnal Tides in Global Surface Pressure Fields', J. Atmos. Sci., 56, 3874-3891