CGIAR-CSI CRU TS 2.1 Global Climate Database - Data Structure

Document Updated: Nov. 2006

VALUE GRIDS, THE CLIMATE VARIABLES AND THEIR DIMENSIONS

Each zipped dataset, either a tile, a continent, or a basin, includes a main workspace which contains a Value Grid, an Info Directory, a Data Directory and a Station Directory.


VALUE GRID

The value grid is in ARCINFO grid format and its geographic attributes are contained in a directory with the same name of the value grid. This grid holds s a uniquevalue for any pixel, which is used as common item to join\relate this grid with the climate attributes tables. The value grid is named using the naming convention cru_<dataset name>_grd.

For Example:
cru_10_20_grd: Values grid for tile 10, 20
cru_Af_grd: Values grid for the Africa continent.

INFO DIRECTORY

The value grid is associated with an INFO directory in which the tabular attributes of the value grid are stored. The value grid (the grid value directory) and its associated info directory must be in the same directory, otherwise the grid can not be used correctly.

DATA DIRECTORY

The CRU dataset includes two types of climate tables for each pixel: the actual CRU climate variables (DATA Directory) and the number of station observations that contribute to the calculation of the climate variable (STN Directory).

The nine climate variables have a variable code used in the file naming convention, an unit and a specific multiplier:
TMP: near-surface mean temperature (degrees Celsius). Multiplier = 0.1
TMN: near-surface minimum temperature (degrees Celsius). Multiplier = 0.1
TMX: near-surface temperature maximum (degrees Celsius). Multiplier = 0.1
DTR: near-surface diurnal temperature range (degrees Celsius). Multiplier = 0.1
PRE: precipitation (mm). Multiplier = 0.1
WET: wet day frequency (days). Multiplier = 0.01
FRS: frost day frequency (days). Multiplier = 0.01
VAP: vapour pressure (hPa). Multiplier = 0.1
CLD: cloud cover (percentage). Multiplier = 0.1

Multipliers are used to increase the precision of the variables values without using decimals. Real or floating values are less efficient in terms of computing time and space compared to integer values. Precipitation values, for instance, have a multiplier equal to 0.1. Values retrieved from the CRU dataset have to be multiplied for 0.1 to retrieve the values in the correct units (a precipitation value equal to 143 in the CRU dataset is equivalent to 14.3 mm.)

The lookup tables that contain climate variable values are characterized by the code ‘data’ within the data directory.
The time series for each climatic variables has been separated in 6 different tables (comma delimited ascii files) with 6 different time ranges (1901-1920, 1921-1940, 1941-1960, 1961-1980, 1981-2000 and 2001-2002).

The data directory contains also an additional lookup table, named ‘coordinates.txt’, with coordinate values (x,y) for each pixel in either geographic (‘x_ll’,’ y_ll’) and CRU (‘x_cru’,’ y_cru’) coordinate system.

STATION DIRECTORY

Information on the changing network of stations (number of stations) that contribute to the value at an individual grid-box are available. While CRU is not able to release the station data (subject to non-disclosure agreements), it provides additional files that record the number of stations for each month within range of a pixel.

The station directory contain station information for the following seven climate variables, which attribute values are described in the DATA directory: TMP, DTR, PRE, WET, FRS, VAP, CLD. Information on the station numbers for the TMX and TMN data values are equal to the station numbers of the DTR variable. Each of these tables express the number of station observation within the pixel without the use of any multiplier.

The tables that contain climate the number of station observations are within the stn directory and are characterized by the code ‘stn’.
The time series for each climatic variables has been separated in 6 different tables (comma delimited ascii files) with 6 different time ranges (1901-1920, 1921-1940, 1941-1960, 1961-1980, 1981-2000 and 2001-2002).

As for the DATA directory, the STN directory contains also an additional lookup table, named ‘coordinates.txt’, with coordinate values (x,y) for each pixel in either geographic (‘x_ll’,’ y_ll’) and CRU (‘x_cru’,’ y_cru’) coordinate system.

NAMING CONVENTION FOR CLIMATE TABLES

<climate variable>_<beginning year>-<ending year>_<data type>.txt

where:
climate variable: climate variable (TMP, TMN, TMX, DTR, PRE, WET, FRS, VAP, CLD)
beginning year: first year of the time series described in the table
last year: last year of the time series described in the table
data type: either variable values (data) or variable station numbers (stn)

Each of the time series defines in its value attribute table climate values for each month for the period 1920-2002. The months are defined using the following convention:

TMP_1901-1920_data = mean monthly temperature values between January 1901 and December 1920
DTR_2001-2002_stn = station numbers used to calculate mean monthly temperature range between January 2001 and December 2002

Within each table, every field describes information for each month between 1901 and 2002 using the following coding:

M<month>Y<Year>

For example:
M11Y1972. month 11 (november), year 1972

Every table has less than 256 columns, therefore it is possible to use these tables across different softwares and convert the tables to dbf, which have a limit of 256 columns. This dataset has been tested in ArcGIS 8/9, Arcview 3x, Excel and Access.

For questions related to the original dataset, please refer to the CRU TS 2.1 web page: http://www.cru.uea.ac.uk/~timm/grid/CRU_TS_2_1.html