Monitoring the Impacts of Disruption Events on Agriculture Through Irrigation Detection with Remote Sensing

Adoption of irrigation even on a small scale requires access to suitable soils, sourcing seeds and fertilizers, meeting crop water requirements, energy and infrastructure needed for pumping water and access to markets. Therefore monitoring irrigation over time provides key insight on the effectiveness of policy decisions, infrastructure successes and more importantly, disruptions caused by events such as conflicts, the COVID pandemic etc. All of these have a direct impact on food security. However most of the irrigation maps including coarse global irrigation cropland layers are created for a certain time period and do not reflect changes in the landscape over time. This study presents a simplified alternative to a dry season irrigation detection methodology developed for the Ethiopian highlands that can be implemented using Google Earth Engine's built in classifiers to create irrigation maps for each year of interest. The original methodology and the proposed simplified approach are compared for a sample region in Oromia to discuss the trade-offs of the simplifications. It is then applied to Southern Tigray across five years from 2019 to 2023 to show how the monitoring technique can be used to visualize the effects of disruptions such as the Tigray War on irrigated agriculture and the recovery efforts after.