Using the CGIAR-CSI CRU 2.1 Dataset with ESRI GIS Software Packages

These datasets have been tested to work with ArcView 3.x, ArcInfo 8/9 and ArcMap 8/9.
In this section we briefly outline how to open and use the CRU dataset with these software packages.

ARCVIEW 3.x

Activate the spatial analyst extension and add to a view the value grid, located in the main workspace directory. Once the value grid is loaded, open its theme table which contains two fields. The first field(“value”), shows an unique value for each pixel and it is used as relate field to retrieve monthly climate values values from the look-up tables in the "DATA" and "STN" directories. The second field is the “count” field, which always equals 1 since each pixel has an unique value. The count field is needed for the grid to be loaded correctly, do not delete this field.

With the table class selected, in the window project, add the tables relative to the climate variables and the time range of interest. You can now join these climate tables to the grid value table once at the time, using the “value” field as common field. Open and make active the climate table, click on the “value” field in the CRU database table to make this field active. Open and make active the value grid table, click on the “value” field in the value grid table to make the field active. Join the tables using the “join” button or the join option from the “Table” pull down menu.

The contents of the value grid table changes to include the joined attributes from the climate tables. Once joined, the value grid can be labeled, queried and symbolized using the attributes of the joined climate tables. The joined attributes are not permanently part of the ArcView table.

You might join several climate tables to the grid value table and query them, but keep in mind that a table can be exported and opened correctly in excel and arcmap only if there are less than 256 columns. Each one of the CRU climate variables has been split in time periods of 20 years for climate tables with less than 256 columns. Therefore if you try to export datasets between different softwares it is necessary to join and export one climate table at the time.

In case the data are used to interact with many tables simultaneously in ArcView using the tabular or graphic interface, it is possible to simply relate the tables using the link option available in ArcView. You can link a value grid table to one or more climate tables based on the common field “value”found in both tables. Open and make active the climate tables, click on the “value” field in the climate tables to make this field active. Open and make active the value grid table, click on the “value” field in the value grid table to make this field active. Link the tables using the “link” button or the link option from the “Table” pull down menu. Repeat this procedure for any climate table that you need to relate to the grid value table.

Unlike joining tables, linking tables simply defines a relationship between two tables, rather than appending the fields of the source table to those in the destination. When tables are linked, neither table is changed - they are just linked to one another. After a Link is performed, selecting pixels through the graphical interface or several records in the value grid table will automatically select the record or records related to it in the loaded and linked CRU climate tables.

ArcMap Version 8/9

Add the value grid contained in the main workspace directory and the monthly climate tables relative to the climate variables and the time period of interest, located in the "DATA" and "STN" directories.

You can join the climate tables to the grid value table once at the time, using the “value” field as common field. Right click on the value grid in the table of contents, and select the “Join..” command from the “Joins and Relates” option. Click the first dropdown arrow and click Join attributes from a table. Click the second dropdown arrow and select the field “value”. Click the third dropdown arrow to choose the climate table to join, and select the field “value” in the fourth dropdown arrow.

The contents of the value grid table changes to include the joined attributes from the climate table. The joined attributes are not permanently part of the value grid table. If you want to permanently save the grid with the joined climate data, export the value grid as a new grid. Once joined, the climate attribute can be queried, retrieved and plot using the query and/or graphical selections on the grid value.

You might join several climate tables to the grid value table and query them, but keep in mind that a table can be exported and opened correctly in excel, arcmap or arcinfo only if there are less than 256 columns. Each one of the CRU climate variables has been split in time periods of 20 years for climate with less than 256 columns. Therefore if you try to export datasets between different softwares it is necessary to join and export one climate table at the time.

In case the climate data are used to interact with many climate tables simultanously in ArcView using the tabular of graphic interface, it is possible to simply relate the tables using the relate option available in Arcmap. You can build a relation between two tables based on the common field (“value”) found in both tables. Right click on the value grid in the table of contents, selecting the “Relate..” command from the “Joins and Relates” option. Click the first dropdown arrow and select the field “value”. Click the second dropdown arrow to choose the CRU climate table to relate, select the field “value” from the third dropdown arrow. Type a name to be used to save this relation. Once you have created a relation, you can access the related records by clicking options in the grid value tables and click the name of the relate you want to access.

Unlike joining tables, relating tables simply defines a relationship between two tables, rather than appending the fields of the source table to those in the destination. After a relate is performed, selecting pixels through the graphical interface or several records in the value grid table will automatically select the record or records related to it in the loaded and related CRU climate tables.

ArcINFO

ArcInfo can import dbf table, unlike txt (comma separated) tables. If you want to import CRU climate tables in ArcInfo, export the climate txt tables to dbf using softwares like Excel or ArcView. Once converted in DBF, Arcinfo can convert the DBF table to INFO table using the DBASEINFO command at the Arc prompt.

DBASEINFO <CRU Climate DBF table - input> <CRU Climate INFO table - output>

The newly created INFO tables can then be joined or related to the value attribute table (VAT) of the value Grid, using the JOINITEM command at the arc prompt. Be aware that in ArcInfo, unlike ArcView or ArcMap, joins are not temporary. We therefore advise to join the climate tables to a copy of the value grid (COPY command at Arc prompt). INFO tables can handle up to almost 4100 columns, while DBFf files can handle only 256 columns. It is possible to join up to 4086 fields (monthly climate values) from the CRU climate table to the value grid table.

JOINITEM <Grid Value Attribute Table> <CRU Climate INFO table> < Grid Value Attribute Table > <”Value”>

RELATE ADD <relate name> < Grid Value Attribute Table > < CRU climate_info_file > <”Value”> <”Value”>

Individual climate grids can be created for each months for a specific CRU variable using AML scripts from a Value grid, which has been joined to CRU climate tables. For instance if the tile 20 20 (“cru_20_20_grd”) has being joined with precipitation monthly values from 1960 to 2002, it is possible to create individual precipitation grids for each month, using a simple AML script:

&do m = 1 &to 12
&do y = 1960 &to 2002
pre_%m%_%y% = cru_20_20_grd.m%m%y%y%
&end
&end