Operational Early Warning System Using SPOT-VEGETATION and TERRA-MODIS to Predict Desert Locust Outbreaks

Abstract

The Desert Locust (Schistocerca gregaria) is an insect whose distribution area extends from North Africa through the Near East to Southwest Asia. During invasion periods, adults form swarms that can fly or be carried by wind over great distances. These swarms can threaten crops located thousands of kilometres from their places of origin. Supported by its member countries, FAO established the “Emergency Prevention System for Transboundary Animal and Plant Pests and Diseases” (EMPRES programme – Desert Locust Component) to minimize the risk of emergencies developing as a result of Desert Locust plagues. EMPRES has created an operational early warning system involving surveying for Locusts in order to find Desert Locust outbreaks as early as possible and prevent them from developing into serious upsurges or plagues. The system includes the use of SPOT-VEGETATION images received on a ten-day basis at FAO through ARTEMIS. The S10 SPOT-VEGETATION NDVI product and single channels (RED-NIR-SWIR) are analysed at the Desert Locust Information Service, FAO HQ for conditions favourable for Desert Locust reproduction and development. The images and analyses are then provided to each of the twenty countries via a FTP site where the Locust Information Officers can have direct access. The analyses aim to identify two sorts of error: i) Commission error (when NDVI indicates that there is vegetation but there is no real vegetation on the ground) ii) Omission error (when NDVI indicates no presence of vegetation but there is vegetation on the ground). To reduce the first type of error, S10 single channels (Red, NIR, and SWIR) are needed and provided to the users. To reduce the second type of error, TERRA-MODIS images are used to detect sparse vegetation not identified with SPOT-VEGETATION. The satellite images (SPOT-VEGETATION and TERRA-MODIS) are now used operationally by some Locust survey teams to direct their surveys toward high-risk regions and allow them to optimize use of available resources. Once in the field, survey teams introduce field observations into a palm-top computer and send the information to the National Locust Unit (LCU) in real-time via radio signal. The information arriving at the centre is then automatically integrated into a GIS system that was developed specifically to manage Locust and environmental data. This system, called the Reconnaissance And Management System of the Environment of Schistocerca (RAMSES), allows Locust Information Officers to orient field surveys geographically, to predict Desert Locust breeding and migration and to develop control strategies in case of emergency. The integration of satellite images within RAMSES also allows the officers to assess the areas covered by vegetation where control operations must be carried out. At the Desert Locust Information Service (DLIS), FAO HQ (Rome, Italy), all of the data collected at national levels are automatically imported into a much larger UNIX-based GIS called SWARMS (Schistocerca WARning Management System) where short and medium-term forecasts are prepared indicating potential locust migrations and areas of breeding.