Towards a crop-specific drought impact assessment tool in Chiredzi District, Zimbabwe, using Sentinel-2 and Landsat-8/9 imagery

Details

Supervisors

Defourny, Pierre ; Bontemps, Sophie

Faculty

Faculté des bioingénieurs

Degree label

Master [120] : bioingénieur en sciences et technologies de l'environnement, à finalité spécialisée

Abstract

The increasing frequency and severity of droughts due to climate change pose significant threats to agricultural production, particularly in regions like Southern Africa. Accurate monitoring and assessment of drought impacts at the crop-specific level are crucial for effective agricultural management and food security. This study proposes a valuable tool, a high-resolution Vegetation Health Index tailored for monitoring drought effects on various crop types within the Chiredzi District of Zimbabwe. Utilizing Sentinel-2 satellite imagery, NDVI and LST time series were calculated to produce a VHI with a spatial resolution of 10 meters per pixel. The VHI was further refined by correcting for seasonal shifts, ensuring that crop development phases were accurately aligned across different years. This high-resolution index was integrated with a crop type classification map, derived using Random Forest classification, enabling precise assessment of drought impacts on individual crops at this fine spatial scale.The results reveal significant variability in drought resilience among different crops, with rain-fed crops like sorghum being particularly vulnerable. The study underscores the importance of high-resolution data and seasonal adjustment in enhancing the accuracy of drought monitoring systems, offering a valuable tool for improving agricultural decision-making in drought-prone regions. The methodology developed here can be adapted to other regions, contributing to the refinement of early warning systems for food security.