Soil moisture from remote sensing to forecast desert locust presence

Abstract

Preventive control of desert locusts is based on monitoring recession areas to detect outbreaks. Remote sensing has been increasingly used in the preventive control strategy. Soil moisture is a major ecological driver of desert locust populations but is still missing in the current imagery toolkit for preventive management. By means of statistical analyses, combining field observations of locust presence/absence and soil moisture estimates at 1 km resolution from a disaggregation algorithm, we assess the potential of soil moisture to help preventive management of desert locust. We observe that a soil moisture dynamics increase of above 0.09 cm3/cm3 for 20 days followed by a decrease of soil moisture may increase the chance to observe locusts 70 days later. We estimate the gains in early warning timing compared to using imagery from vegetation to be 3 weeks. We demonstrate that forecasting errors may be reduced by the combination of several types of indicators such as soil moisture and vegetation index in a common statistical model forecasting locust presence. Policy implications. Soil moisture estimates at 1 km resolution should be used to plan desert locust surveys in preventive management. When soil moisture increases in a dry area of potential habitat for the desert locust, field surveys should be conducted two months later to evaluate the need of further preventive actions. Remote sensing estimates of soil moisture could also be used for other applications of integrated pest management.

Cartographie des zones de reproduction et de grégarisation du criquet pèlerin au Tchad

“…Le dispositif de prévention a été amélioré ces dernières années à l'aide de nouveaux outils (Cressman, 2008 ;Zhang et al, 2019) comme la télédétection spatiale (Latchininsky et al, 2016). Il est maintenant possible de collecter en temps réel des informations clés sur les conditions écologiques favorables à la reproduction, de les traiter très rapidement et d'envoyer les équipes de prospection dans les zones ayant une forte probabilité d'abriter des populations de criquets pèlerins (Piou et al, 2017(Piou et al, , 2019. Un premier atlas des zones de reproduction (Popov, 1997) a été publié par l'Organisation des Nations Unies pour l'alimentation et l'agriculture (FAO).…”

Application of Remote Sensing Data for Locust Research and Management—A Review

“…The observed accumulation of studies is related to locust outbreaks rather than technological advances and availability of remote sensing data. However, recent analysis related to soil moisture [ 51 , 52 , 53 , 54 ] as well as ecological niche modelling [ 5 , 55 , 56 ] based on several data sources were the focus of investigation and showed promising results.…”

“…Using a set of SDMs and climate variables, the authors concluded strong niche conservatism between both subspecies. Piou et al [ 54 ] investigated temporal development for NDVI, soil moisture, rainfall and land surface temperature around survey points of desert locust presence in recession areas. The authors applied statistical analysis for all variables separately to assess their individual potential to explain and forecast desert locust presence.…”

On the relative role of climate change and management in the current desert locust outbreak in East Africa

“…During the 20th century, research and international cooperation resulted, under the aegis of the Food and Agriculture Organization (Lecoq, 2003), in a concerted management effort based on identifying local hotspots where favourable conditions occur (e.g. Piou et al, 2019), to control potential swarm formation before they transform into a large-scale unmanageable threat (Figure 1a). This approach resulted in a considerable decrease in the number and intensity of locust outbreaks and plagues (Figure 1b).…”