Last update 13 February 2002 Air Pressure Dataset Information The following is a set of notes compiled from information supplied by various people on the availability of different air pressure datasets and on general background. Information was kindly provided by the following people or organisations or abstracted from the following sources: (a) Roy Jenne (NCAR) (b) Peter Dexter (WMO) (c) UK Met Office/University of East Anglia Climatic Research Unit (d) National Climatic Data Center (for NCDC and CDIAC information) (e) NOAA/NESDIS (f) the World Climate Disc leaflet (g) the World WeatherDisc leaflet (h) US Navy Marine CD-ROM Climatic Atlas information and (i) the WCRP Infoclima report (j) advertisement for a book by Allan et al. (k) UK Met Office (Rob Allan) (l) NOAA-CIRES Climate Diagnostics Center (m) Joint Institute for the Study of the Atmosphere and Ocean (n) KNMI Climate Explorer Analysts of the PSMSL data set who need monthly mean air pressure information would probably find (c) and (k-n) the most useful. Some of the other sets of information may be not be up-to-date. Copies of these datasets should be obtained from the original sources. The PSMSL cannot offer to copy them. Dr. Rob Allan (UK Met Office) is thanked for much of the recent information shown. ============================================================================ (a) Information from Roy Jenne (NCAR) (jenne@ncar.ucar.edu) on datasets available from NCAR and other places (latest version 23 August 1995). 1. Summary of Available Sea Level Pressure Grid Data (rev 23 Aug 1995) This text summarizes the basic SLP grid point data that are available. Other datasets have the basic observations from ocean areas, land areas, and ice caps. The text now also includes a little information about station data. 1.1 Northern Hemisphere Daily (and Monthly) SLP, 1899-on The data came from several sources including the U.S. Navy. The data are based on U.S. weather service maps until around 1960. In the last 20 years we have tried to select sources that avoid most problems. A list of sources used is available. The basic monthly grids are derived from the daily grids, but there is a tape by Trenberth where a few corrections have been applied to the basic monthly grids. 2. NMC Daily Global Analyses SLP is okay from the start date 1 Jul 1976. Wind grids are bad in the tropics until Sep 1978. 3. Southern Hemisphere Daily SLP Grids for: a. IGY, Jun 1957 - Dec 1958; all Southern Hemisphere data is from South Africa. The quality is very good. b. 60-210E, 1957-1978, from New Zealand. The data were based on NZ operational analyses, and then adjusted to account for information in later data taken up to 24 hours beyond analysis time. We think that the quality should be almost as good as for the IGY grids, except on the Western edge. c. Southern Hemisphere, Apr 1972-on, from Australia. Australia used satellite cloud pictures to help position fronts. d. Southern Hemisphere, 1974-on, from the U.S. Navy e. Southern Hemisphere, Jan 1950 - May 1957, daily grids from South Africa. The quality is poor in some areas, mainly the big ocean area west of Chile and south of Easter Island. 4. Summary for S. Hemisphere Daily Grids: a. For Jun 1957-95, there are daily grids for the NZ sector (60-210E). b. The whole S. Hemisphere has grids for Jun 1957 - Dec 1958 and Apr 1972-on. S. Africa has map analyses through 1963 that could be digitized to help fill gaps. c. There is now a gap for two-thirds of the hemisphere for Jan 1959 - Mar 1972. d. There are grids also for Jan 50 - May 1957. 5. Data for Tropical Storms Tropical storms are often very intense. Because of grid resolution and other problems, they often have not been represented well in the analyses. NCAR has a dataset (and text) giving tropical storm data. The central pressure, and often winds, are given each 6 or 12 hours. The dataset spans many years. 6. COADS Surface Marine Observations Observations from ships and buoys are sent on GTS each few hours. They are also recorded on logbooks on ships. The logbooks are digitized. A big project called COADS brings all this data together into one dataset. These reports (pressure, temperature, wind, SST, etc.) are being used for reanalysis. Also, year-month statistics are prepared directly from the ship data. 7. COADS Yr-Mo Statistics for SLP, winds, temperature, etc. Data is for 2-degree boxes, for many years. When analyses exist, they often should be better than these statistics. The number of samples in a box is often too small for a stable mean. 8. Northern Hemisphere Daily SLP Data from the USSR (in 1987) This data is for a 5 x 10 grid, 1880-1979, and is on 4 tapes. Note: Consider the monthly SLP data that started 1880; I think that it has become confused whether the original source for earlier years was the UK or Russia. In any case, the 1899-on data in item No. 1 is different,and it is from U.S. sources. 9. Reanalysis Data for 1957-on There is an NMC/NCAR project to reanalyze the whole atmosphere (and ocean) each 6 hours, from 1957-on. See Bull AMS, Dec 1991 (and late 1995). This would produce atmospheric analyses, SLP, 10m winds, and ocean flux terms (radiation, heat, etc.) NCAR is working hard to prepare mountains of data. The production phase started Jun 1994. By late Aug 1995, data for 1982-93 (12 years) had been completed. 10. Climatology NCAR has good monthly world SLP means on 5-degree grids based on all data up to about 1960 or 1965. These data should be good, even in parts of the Southern oceans, where the data is thin to define yr-mo means. For the S. Hemisphere, these are based on "Climate of the Upper Air, Part 1 - Southern Hemisphere," Taljaard et al., 1969. Navair 50-1C-55. A good summary is in "A Selected Climatology of the Southern Hemisphere: computer methods and data availability." NCAR TN/STR-92, Jenne et al., 1974. 11. Southern Hemisphere Monthly SLP Phil Jones wrote a paper which is published in the Journal of Climate, Dec 1988, where he used monthly gridded data to determine patterns of SLP. This information was then used with the available older station data to try to determine the best monthly analyses (S. Hemisphere) for the whole period back to 1957. Another paper went back to 1911, but this is on shakey ground (International Journal of Climate, Vol II, P585-607, 1991). 12. Conservation of Atmosphere Mass We need some more comparisons between the various SLP datasets. We should prepare RMS differences of SLP between some key datasets over remote areas. The total mass of dry air in the atmosphere should be very close to constant. Global water vapor can vary some from month to month, but it isn't a large term. Global analyses of SLP can be evaluated to see if mass is conserved (a calculated surface pressure is actually used). Trenberth did some work on this (JGR, D12, Dec 20, 1987), and (JGR Vol 99, D11, Nov 20, 1994). 13. Monthly pressure data from stations NCAR has a tape with year-monthly data for around 2000 stations, global coverage. The elements given are typically temperature, precipitation, station pressure, and sea level pressure. 14. How to obtain the data Nearly all of the above data is in the NCAR archives. 15. Sea Level Pressure and Surface Pressure A station pressure is always what is measured and that is some height above sea level. To calculate a sea level pressure, some assumption must be made for the temperature of the fake air column between the surface and sea level. Over the oceans and for low elevations, this is no problem. For high elevations, and especially for sharp cold air surface inversions, it is a problem. Rules were developed so that land stations can calculate a fairly reasonable SLP. Therefore, the best gridded SLP has often been obtained by using the reported station SLP in the analyses. For the past 15 years or so, the main analysis schemes do not use SLP from most land stations; they define a model surface elevation and analyze for surface pressure at that height. This is actually the best thing to do because it involves using information that is measured rather than guesses about fake air temperatures underground. The only problem has been that when the centers do derive a sea level pressure, they often have not taken enough care to estimate a reasonable mean temperature. They could do this better than an observer at a station can, but often don't Another possibility would be to derive a SLP that has horizontal pressure gradients consistent with those at the elevated surface. ================================================================================ (b) Information on measurement methods and data availability from Peter Dexter (dexter@www.wmo.ch) (World Meteorological Organistion) in October 1992. Checked in August 1995 to be still basically correct: (1) Current measurement practices and accuracy requirements are detailed in WMO No.8 'Guide to meteorological instruments and methods of observation' and WMO No.717 'Report of the tenth session of the WMO commission for instruments and methods of observation'. The latter contains information in particular on the accuracy requirements for automatic weather stations. (2) Atmospheric pressure measurements are made at all stations in the basic synoptic network. There are thousands of these worldwide, and a complete list is given in WMO No.9 Volume A 'Observing stations'. If you want information for particular countries it may be best to contact the national Meteorological Services directly, rather than wade through the complete listings. Contacts for all national Services may be got from WMO No.2 'Meteorological services of the world'. (3) All meteorological data collected nationally from terrestrial stations are also archived nationally. Data archival procedures and access policies (including on charging) may vary from country to country. The best initial approach is therefore to contact the respective national Service for advice and elaboration. Data from at least the basic synoptic network stations are also stored in global archives through the ICSU-WDC network. The best first contact with this network would be WDC-A for Meteorology, which is the National Climatic Data Center, Asheville, NC, USA. The full contact address for this will also be in WMO No.2. The above publications are available from WMO but all are sale publications (i.e. not free). ================================================================================ (c) Grid-Point Pressure Data for the Northern Hemisphere (August 1999) The UK Met Office and University of East Anglia provide monthly mean air pressures on a 5 x 10 (lat x lon) degree grid for 1873 onwards for north of 15 North. See http://www.cru.uea.ac.uk/cru/data/pressure.htm for details. ================================================================================ (d) Information from the National Climatic Data Center (NCDC) (updated February 2002) The NCDC (http://www.ncdc.noaa.gov) and the World Data Center for Meteorology (http://lwf.ncdc.noaa.gov/oa/wmo/wdcamet.html) can provide a number of relevant products. There is the Monthly Climatic Data for the World (MCDW TD 9645) database which is derived from a combination of efforts. This dataset contains monthly means of air pressures etc. at meteorological stations (often airports). Note that the MCDW database is not incorporated into a dataset called the Global Historical Climatological network (GHCN) which is available via WDCM web pages although with restrictions on distribution. That data base contains monthly temperature (mean), total precipitation, sea-level pressure, and station-level pressure for thousands of stations. One can access the GHCN version 1, which has the pressure information up to 1990 at Oak Ridge National Laboratory. This is provided free of charge from CDIAC as the NDP-041 dataset: http://cdiac.esd.ornl.gov/epubs/ndp/ndp041/ndp041.html This leads to a written report, the GHCN V1 digital data sets, and FORTRAN and SASTM data retrieval programs. Because of the size of the data base, the NDP is not available on floppy diskette. Version 2 of the data set does not at the time of writing include air pressure data. ================================================================================ (e) Information from Christine Young at NOAA/NESDIS (August 1995) The two volumes: NOAA Atlas NESDIS 6 Atlas of Surface Marine Data 1994 Volume 1: Algorithms and Procedures and NOAA Atlas NESDIS 7 Atlas of Surface Marine Data 1994 Volume 2: Anomalies of Directly Observed Quantities describe the reanalysis of monthly mean fields, including air pressures, over the global ocean from the COADS datasets, spanning the period January 1945 to December 1989. Sea level pressures are calculated from individual observations taken by ships, buoys, and other marine platforms analysed on a 1x1 degree grid over the global oceans (not over land). A climatology analysis exists for each month of the year covering the period 1945-1989. Besides these 12 climatology fields, there are 540 (12 months x 45 years) monthly anomalies (analyzed departures from the monthly climatology) also analysed on a 1x1 degree grid. The analysis consists of an objective analysis scheme (Barnes) along with linear and non-linear filters to interpolate/extrapolate to grid points missing information and to remove noise. Unanalyzed fields are also available. This sea level pressure analysis is just a small part of a larger set based on the COADS marine observations. Other fields include all directly observed quantities (sst, sat, specific humidity, cloudiness, winds), heat, momentum, and fresh water fluxes (wind stress, latent and sensible heat flux, long and short wave radiation, constrained net heat flux, precipitation, evaporation, buoyancy flux), and other useful derived quantities (air density, vapor pressure, sst-sat, qs-q, cube of friction velocity, etc.). All quantities are analyzed similarly and cover the same period. Most analyzed quantities will be available on a 3 CD-ROM set. Both analyzed and unanalyzed fields will be available on exabyte (8mm) tape. A multiple volume atlas will also be available to explain all methods/computations and depict the analysis in a graphical manner. For more information: Email: services@nodc.noaa.gov NODC Home Page: http://www.nodc.noaa.gov/ ================================================================================ (f) The World Climate Disc (information from October 1992, updated August 1995). The World Climate Disc contains global climatic change data on CD-ROM and has easy-to-use software for creating maps etc. It contains: - Monthly mean surface air temperature data for over 3100 stations from 1854 to 1990. - 5 deg x 5 deg gridded monthly mean temperatures from 1854 to 1990. - Monthly total precipitation data for over 7300 stations from the early eighteenth century to 1990. - 5 deg x 10 deg gridded mean sea-level pressure data for the northern hemisphere from 1873 to 1990. - 5 min x 5 min gridded mean height data for the whole world. The disk has been produced by the Climatic Research Unit, University of East Anglia with the air pressure data compiled by the Hadley Centre (U.K. Met.Office). The price is 595 UK pounds. Orders can be made to: Chadwyck-Healey Ltd., Cambridge Place, Cambridge CB2 1NR, U.K. Tel: 0223 - 311479. Fax: 0223 - 66440. In North America: Chadwyck-Healey Inc., 1101 King Street, Alexandria, VA 22314, USA. Tel: (703) 683 - 4890. From Canada Tel: (800) 535 - 0228. In France: Chadwyck-Healey France S.A., 3 rue de Marivaux, 75002 Paris. Tel: 1 42-86-80-20 Fax: 1 42-61-33-87. In Spain: Chadwyck-Healey Espana S.L., Calle Recoletos 11, 28001 Madrid Tel: 575 23 14 Fax: 575 98 85 From a conversation from Dr.Phil.Jones (CRU/UEA) in August 1995 it seems that an updated version of this disk is not planned but Dr.Jones will be pleased to discuss the availability of new data additions as appropriate. ================================================================================ (g) The World Weather Disc (information from November 1995). CDROMs for both DOS and Macintosh systems containing 17 datasets that form a meteorological database that describes the climate of the past few hundred years. It has been derived from the archives of NCAR and the National Climate Data Centre. It includes: (1) World monthly surface station climatology. Time series of monthly mean temperature, precipitation and sea level and/or station pressure for 1000's of stations around the world. Some records go back to the 1700's. (2) Worldwide airfield summaries. Climatological summaries with a wide range of parameters for 100's of airports. (3) daily weather at hundreds of US stations. (4) data sets such as Climatology of the US No.20, Local Climatological Data (LCD), US Climatic Division Data, US monthly normals, and worldwide airfiled summaries. Parameters include temperature, precipitation, heating/cooling degree days, freeze occurrence, wind, sunshine, lightning, tornadoes, tropical storms and thunderstorms. In the UK, this dataset can be inspected by contacting Dr.Gray at the Rutherford Appleton Laboratory (Tel: 01235 446745). General information from WeatherDisc Associates Inc., 4584 NE 89th, Seattle, Washington 98115, USA. Tel: (206) 524-4314. ================================================================================ (h) US Navy Marine CD-ROM Climatic Atlas (information from October 1992). The following was obtained from Don Collins, NODC User Services (services@nodc.noaa.gov): A new CD-ROM entitled US Navy Marine Climatic Atlas of the World Version 1.0 has been released. The CDROM was produced by the National Climatic Data Center, Asheville under a project funded by the Naval Oceanography Command Detachment, Asheville, NC. The CDROM includes analysis and display software for monthly climatological averages of atmospheric and oceanographic data observed within 1 and 5 degree grid areas covering the global marine environment. Available elements include air and sea temperature, dew point temperature, scalar wind speed, sea level pressure, wave height, wind and current roses, and probability of superstructure icing and gale force winds. The summary stattistics are derived from NCDC's marine database covering the period 1854-1969. The cost is 61 US dollars from National Climatic Data Center, Federal Building, Asheville NC 28801 or credit card calls to 704-259-0682. Fax orders by credit card to 704-259-0876. ================================================================================ (i) The WCRP Infoclima (Catalogue of Climate System Data Sets) Report The latest edition of this report (WCDP-5 or WMO/TD-No.293) dates from April 1989, although there is a supplement dated March 1992 not relevant to this note. A new version of Infoclima is expected by early 1993. Copies can be obtained from the World Meteorological Organisation (WMO), 41 Giuseppe-Motta, Case Postale No.2300, CH-1211 Geneva, Switzerland (Telex 414199). This report comments on the availability of several air pressure datasets, including the omnipresent 5 deg x 10 deg monthly means 1899 on for the northern hemisphere, 1972 on for the southern hemisphere (see pages 163- 165). ================================================================================ (j) "El Nino Southern Oscillation and Climatic Variability" by Rob Allan, Janette Lindesay and David Parker. CSIRO Publishing. Sept/Oct 1996. 408pp. ISBN: 0 643 05803 6. This book includes a cdrom with global monthly gridded fields of mean sea level pressure and sea surface temperature anomalies since 1871. Contact: sales@publish.csiro.au or: http://www.dar.csiro.au/pub/info/publications/allan.htm ================================================================================ (k) UK Met Office Dataset of Gridded Mean Sea Level Pressure (February 2002). This data set called 'GMSLP2.1f' consists of a 5 * 5 degree global grid of monthly mean air pressures for 1949-1994. It will be extended backwards eventually to 1871 but will not be extended forwards in the future as another data set will take its place. As of February 2002, Rob Allan is working on these new data sets with the ultimate aim of extending them back to about 1850, and also on a data set of station rather than gridded data. Contact Rob Allan (rob.allan@metoffice.com). See also http://www.met-office.gov.uk/sec5/CR_div/index_climate.html ================================================================================ (l) NOAA-CIRES Climate Diagnostics Center (February 2002) http://www.cdc.noaa.gov/ provides a link to the COADS web site: http://www.cdc.noaa.gov/coads/ which provides ordering information and a link to, for example, http://www.cdc.noaa.gov/coads/r1c.html for documentation. COADS data consist of 2 deg square monthly summaries of many variables (1 deg for later data). Some information goes back to the late 1700s, with more from the 1850s. The same CDC website also provides a link to the NCEP/NCAR Reanalysis Project to provide reanlysed met data (including air pressures) as 6 hourly, daily and monthly products for 1948 onwards. http://www.cdc.noaa.gov/PublicData/web_tools.html allows one to manipulate, composite and correlate various data sets. ================================================================================ (m) Joint Institute for the Study of the Atmosphere and Ocean (February 2002) http://tao.atmos.washington.edu/data_sets/ provides an extensive range of data sets and web links, including versions of COADS and NCEP products and climate indices. ================================================================================ (n) KNMI Climate Explorer (February 2002) http://climexp.knmi.nl/ similarly allows for an extensive range of data sets and web links to many of the products described above. (Simple registration is needed for this site.)