The Desert Locust Monitoring Service: a new Destination Earth service for Environmental Pest prediction

Desert locusts are known as the world’s most destructive migratory pest. A single swarm can count 80 million locusts, traveling up to 150 km daily and consuming the same amount of food as 35.000 people per day. They are cause of major mid-to-long-term impacts on the economy, quality of life and the environment. Climate change is amplifying the occurrence of such pests: the increase of extreme events such as cyclones creates ideal conditions for locust breeding.

The Desert Locust Monitoring Service (DLMS) is the new system for preventing upsurges of desert locusts across Eastern Africa, Southwest Asia and northwest of India. It leverages the power of satellite observations, model outputs, in-situ measurements and advanced AI techniques to monitor the locust threat, help mitigate crop damage, and safeguard essential food supplies.

Onboarded on the Destination Earth (DestinE) platform, the service consists of two layers: the Early-Stage Locust Appearance and the Locust Swarm Migration. The first layer relies on a customized Maxent model, a statistical approach widely used in species distribution modeling (SDM): it allows identifying areas with favorable environmental conditions for desert locust breeding. There are two available modes: the Safe Mode makes use of 50 days of climate data from ERA5-Land (soil water content, precipitation, and temperature) and normalized difference vegetation index (NDVI) from Sentinel-3 to generate daily probability forecasts; the Experimental Mode extends forecasts to two days thanks to the advanced projections of DestinE’s Weather Extremes Digital Twin. The variables used are: skin temperature, total precipitation and runoff.
The second layer forecasts adult locust swarm migration under biological and climatic conditions, accounting for their rapid and unpredictable long-distance movements. It takes as primary input the Early-Stage layer output, which provides initial location predictions under specific environmental conditions. Environmental variables, including the Leaf Area Index (LAI) from Copernicus, as well as wind velocity components and temperature from DestinE’s Climate Change Adaptation Digital Twin, inform the model of conditions that may trigger migration events. Swarm behavior is then represented using a stochastic model, which simulates an environment-biased random movement on a 2D lattice, generating batches of diverse potential scenarios. In this framework, locusts move based on environmental cues, including climate conditions and the availability of resources, such as vegetation. Finally, the model performs a statistical analysis across all generated scenarios to produce output maps estimating future locations of adult locusts and the size of their swarms.
Both Maxent and stochastic models were trained using a presence-only dataset provided by FAO’s Locust Watch. Prediction results have been validated with FAO data on desert locust activity and independent data provided by the International Center of Insect Physiology and Ecology (ICIPE).

Enabled by Destination Earth’s cutting-edge modeling and data infrastructure, the DLMS offers high-quality insights to anticipate risks, mitigate impacts and support the protection of crops, communities, and ecosystems in the most affected regions. Forecasting maps are available to all DestinE registered users, with some features reserved to users with upgraded access.