GC094-04
Identifying Opportunities to Electrify Diesel-Powered Irrigation Pumps Using Remote Sensing Data in Ethiopia
Abstract:
In this work, we propose a novel method to explore whether off-grid diesel-powered irrigation activity in Ethiopia can be identified using satellite observations of pollution data. The recently launched TROPOMI instrument on the Sentinel-5 Precursor satellite provides air quality and climate-related atmospheric constituents data at unprecedented spatial and temporal resolution. To achieve this, we leverage the TROPOMI NO2 and CO datasets together with other datasets on population, crop production, and atmospheric variables. Existing land use regression models are used to estimate in-situ measurements of NO2 and CO. Using machine learning techniques, we then investigate spatial and seasonal patterns of elevated NO2 and CO levels and their correlation with irrigation seasons, low population densities, and high crop yields. Follow up work involves validating this technique with diesel irrigation pump location data collected via surveys. At scale, this technique can guide electrification of diesel irrigation pumps across Ethiopia, improving environmental and financial sustainability for farmers and utilities.