High resolution Mapping of Agricultural Water Productivity using SEBAL in a cultivated African Catchment, Tanzania

This study aims at high resolution (10m) mapping and evaluation of the spatial variability of biomass, yield, ET and WPET in the Makanya river catchment using the automated Surface Energy Balance Algorithm for Land (pySEBAL) with SENTINEL-2 and LANDSAT-8 images, local land use map and locally calibrated leaf area index (LAI) inputs. A coupled phenological variability and supervised classification approach on high resolution images generated a high accuracy LULC layer which was used to map the WPET in the agricultural lands. The pySEBAL results were evaluated in view of local information on crop yields, water allocation and agricultural management practices in the different agro-ecological zones within the catchment. Calibration of high-resolution satellite LAI generated products with error estimates within acceptable levels of uncertainty. The simulated crop yields were in agreement with reported crop yields. The results showed relatively high WPET in the highlands and low WPET in the midland and lowland areas of the catchment. The latter was attributed to high transmission losses, low irrigation efficiencies, poor agricultural practices and pest/disease attack. When applying SEBAL in African cultivated catchments, it is highly recommended to use SENTINEL-2 data in addition to LANDSAT-8, and to use local information, especially for the ground truthing of land use maps, phenology, crop practices and crop yields.
Source: Physics and Chemistry of the Earth, Parts ABC - Category: Science Source Type: research