| ingrid | The Postscript-based software on which the Data Library is built. | 
| CPT | Climate Predictability Tool More information | 
| ferret | Interactive computer visualization and analysis software. More information | 
| GrADS | Grid Analysis and Display System More information | 
| matlab | Data analysis and visualization software. More information | 
| NCL | NCAR Command Language More information | 
| WinDisp | A public domain software package for the display and analysis of satellite images, maps and associated databases, with an emphasis on early warning for food security. More information | 
| Full Information Formats These files contain all of the available metadata. | |
| OPeNDAP | A system which downloads data directly to software, such as matlab, Ferret, GrADS, etc. Specific instructions are available in the table above. Note: OPeNDAP was formerly known as DODS (Distributed Oceanographic Data System). More Information | 
| netCDF (network Common Data Form) | A commonly supported self-describing data format. More Information | 
| Partial Information Formats These files contain only some of the available metadata. | |
| Columnar Table | A table with separate columns of numbers for each independent variable (i.e., grids) and for the data. This columnar table is available in a variety of file and column formats here. Note: By setting the helper for mime-type text/tab-separated-values, you can have this loaded directly into your favorite spreadsheet. This file will be approximately 1130136 bytes, with 3 columns of 94178 numbers. | 
| 2-Dimensional Tab-Separated Tables | Tab-separated-values (tsv) file with information about the independent variables (i.e., grids). The list to the left allows you to specify the format of the table. Note: The variable running across the top of the table (identifing columns) is listed first and the variable running down the side of the table (identifing rows) is listed second. | 
| GIS-Compatible Formats There are three GIS-compatible formats available. | |
| 2-Dimensional Table | A 2-dimensional ascii file that includes an ArcInfo Header. | 
| IDA Image | File(s) in the Image Display and Analysis format. Typically used with WinDisp. | 
| LAN Image | File(s) in the ERDAS LAN format. Typically used with various GIS programs, including ArcView and HealthMapper. | 
| GeoTIFF Image | File in GeoTIFF format. Typically used with various GIS programs, including ArcView and ENVI. | 
| Data Only Formats These files contain just the data without any of the available metadata. | |
| Binary direct access | A big-endian, ieee single-precision file in floating-point format. Also known as a binary random access file. This is a random-access file; it is purely data with no record-structuring information. The data is structured to correspond to the independent variables (i.e., grids) in X Y order, with the first grid varying the fastest. | 
| DEC ALPHA direct access | Same as the binary random/direct access format above except that it is byte-swapped for DEC ALPHA's and PC's (little-endian). | 
| Binary FORTRAN sequential access | A big-endian, ieee, single-precision file in floating-point format. This is a sequential-access file with each record containing all the X Y points. It must be read using FORTRAN sequential access. There is only one record for the data you have selected. | 
| DEC ALPHA sequential access | Same as the binary sequential access format above except that it is byte-swapped for DEC ALPHA's and PC's (little-endian). | 
| Text with tab-separated-values | Text file where data values corresponding to different X
 are separated by tabs and data values corresponding to different
 Y are on different lines.  This is readable by most programs, including spreadsheets, but will be about four times larger than the binary or netCDF/HDF files noted above. As an alternative, you can access a tab-separated file with the following independent variables (i.e., grids) grouped on each line: X Y | X Y | 
 | 
| Text | Text file where data is arranged in chunks of X Y. There are five values per line and each chunk starts on a new line. This will be about four times larger than the binary or netCDF/HDF files. |