South Africa's climate is characterised by severe spatial and temporal fluctuations in rainfall. This affects issues relating to food security, crop production and general agricultural economics. Policy and decision-makers are often required to respond to drought and other disasters without having sound scientific information available.

In the light of this, the ARC-Institute for Soil, Climate and Water has developed the Umlindi Information System in the form of a website. The system was developed with financial aid from the National Department of Agriculture and attempts to inform decision-makers of the current drought condition, fire risk and vegetation condition. Umlindi is the Zulu word for 'the watchman'.

This information is based on interpreted National Oceanic and Atmospheric Administration (NOAA) satellite and climate data. The data is processed in ESRI's ArcView 3.2 Spatial Analyst and Chips for Windows to derive the following four products. The procedures are fully automated with Avenue scripts.

* Fire: The sites where active fires occurred in a 10-day period. The active fires are derived from the mid-infrared band using Chips. The image derived from Chips for Windows is then displayed according to the required legend in ArcView and exported to the website. The ARC-Institute for Soil, Climate and Water is the southern African node for the World Fire Web. Active fires are detected on a daily basis and sent to an intern server that is available at www.gvm.sai.jrc.it

* Rainfall: Rainfall for a 10-day period interpolated between the available station data. ArcView Spatial Analyst is used to interpolate the rainfall with the following procedure:

* Calculate percentages by dividing the rainfall value with the long-term average for each station.

* Use inverse distance waiting (IDW) to interpolate these percentages.

* Multiply the percentage map with the long-term average map.

The long-term average rainfall surfaces are very accurate. Attributes such as altitude, slope, aspect, distance to sea, position in local terrain, rain shadow effects, etc were used to create these maps.

* Vegetation activity: Processed satellite image to show the vegetation activity. The normalised difference vegetation index (NDVI) images are derived from the red and mid-infrared bands of NOAA images using Chips. The image derived from Chips for Windows is then displayed according to the required legend with ArcView and exported to the website.

* Crop growth and drought: The water satisfaction index for a standard maize variety. It indicates whether this crop has received sufficient water over the season up to the present, or if it has suffered from insufficient water. The water satisfaction index is calculated with ArcView Spatial Analyst from evapo-transpiration and rainfall data.

Users of the website are required to select a province, date and type of query to allow the system to provide information accurately. The various types of queries comprise fire, rainfall, vegetation activity and crop growth and drought. Similar systems are planned for all SADC countries in the near future.

The system went operational in August 2000. There is a delay of five days to process the information for each ten-day period. All software was supplied by GIMS.

GIMS, (011) 315 0390