|
||||||
|
|
||||||
  |
 |
|
|
 |
|
|
|
An analysis of the overlap in station coverage for various precipitation data sets was performed between June 2003 and February 2004 by B. Raup. This analysis was undertaken to provide information that would guide the acquisition of data to fill gaps. The work was funded by the NOAA Arctic Research Program. See NOAA@NSIDC's SEARCH and IPY Activities for more information.
Throughout this analysis, data sets are referred to by a letter and number designation rather than by a title. These NSIDC reference designations are linked to data set title and source in a table. See the Precipitation Data Set Inventory, a .doc file. This table also contains information about bias corrections and coverage for the featured data sets. The table is not meant to be an exhaustive inventory, but rather a summary of the data sets most likely to be important for SEARCH investigations.
Below is a map of station locations for precipitation data sets in the inventory table. (Note that in the map, the C5 dots in Alaska overlay C6 dots.) See the Global Daily Climatology Network for a map of stations from W5.
A first step was to create station location listings in a uniform format. Overlap could then be evaluated using station ID and/or proximity.
These lists have the stations from the collections listed in a uniform format. Each item below is a link to a text file contains information on each station in the collection, with one line per station. The format of the line is:
lon lat NSIDC_reference country station_num station_name year-range
where the fields are separated by TABS. Note that tabs are not always visible when viewing the files in an editor.
(Note that R5 data are from drifting station. Where data are from the same station but locations are different, data are listed as if they were different stations. Also, stations in collection R4 have no station numbers.)
The output below lists duplicated station numbers, each immediately followed by lines listing the data set designation and the relevant entry from that data set. An example:
20049: R2: [57.97, 81.80, 1891-1999, "Rudolf island"] R7: [57.97, 81.80, 1941-1990, "OSTROV RUDOLFA"]
In the above, station 20049 is duplicated in data sets R2 and R7. Data in brackets are [lon, lat, date-range, name].
At the end are listed the number of duplicated stations in each combination of data sets. For example,
R2R2: 5 (34561 35188 35700 36003 38836 )
indicates that there are five stations that are duplicated exactly twice in R2 itself (station numbers listed after). A line such as:
C1C3C3: 1 (6139525 )
means that one station (number 6139525) appears once in the C1 list and twice in the C3 list.
Note that some of the station numbers are followed by an asterisk in the original data set lists; stations with and without asterisks were considered to be the same in this analysis.
(At the very end of the text files are the results of a search for stations in a set of newly acquired data files with names list-52_stn_numbers.txt and ist_51_stn_numbers.txt. This should be ignored, as it is included only for completeness.)
In this analysis, stations which are closer to each other than an input parameter are reported. The input consisted of all stations from all data collections that are north of 45 degrees latitude. The distance parameter was 0.1 degrees, or about 11 km in the north-south direction and about 11 km times cosine(latitude) in the east-west direction. The output is a 7 MB file, and consists of groups of near-by stations, separated by lines containing '--'. Station names often vary in spelling or construction from data set to data set, even for the same station data. Therefore this analysis based on proximity is needed to catch overlap when discrepancies in a station name exist. Station locations can vary too; and the distance parameter used here was chosen after some experimentation.
The data set designated R7 was added to and published after this analysis (see Daily Precipitation Sums at Coastal and Island Russian Arctic Stations, 1940-1990. Some material, such as data plots, a
-- Last modified 5 September 2006 by F. Fetterer, who reformatted and edited the html files provided by B. Raup.
