Monitoring the plant water content in senescent maize at parcel level using multi-frequency SAR data : a case study in the Taiyuan Basin, China

Details

Supervisors

Defourny, Pierre ; Bouchat, Jean

Faculty

Faculté des bioingénieurs

Degree label

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

Abstract

Maize is one of the most widely grown crops globally, playing a pivotal role in agricultural economies and food security. Monitoring the water and dry matter content of maize crops as they reach their final developmental stages is important for ensuring optimal crop yields. This master thesis explores the potential of SAR data, specifically from TerraSAR-X and Sentinel-1 satellites, in assessing water content and biomass dynamics in senescent maize plants by comparing it with in situ-data from the Taiyuan Basin, China. With support from Chinese partners from the Taiyuan University of Technology, field data on senescent maize was collected in the Taiyuan Basin in August and September 2023. Biophysiological parameters including wet and dry biomass and water content were measured on five selected parcels. These efforts aimed to gather comprehensive in-situ data on the final developmental stages of maize crops. A temporal analysis of biomass variables was performed to identify discernible patterns and trends in the maturation process of maize. These patterns were then integrated with concurrent SAR data to comprehensively understand the relationships between in-situ observations and remote sensing data. A linear and non-linear correlation analysis between SAR-derived metrics and field measurements were conducted to quantify their intrinsic relationships. Although the study was not able to establish close correlation between in-situ data and the SAR-derived metrics, it was able to present some unique challenges linked to obtaining in-situ data in the Taiyuan Basin in China, namely the small size of the selected parcels which are often composed of multiple narrow fields and farmed by different owners, leading to high parcel heterogeneity. These findings highlight the complexities of remote sensing in heterogeneous agricultural landscapes and underscore the need for more refined methods and larger, more uniform study areas. Future research could build on this work to develop more robust techniques for integrating SAR data with ground measurements, ultimately improving the precision of agricultural monitoring and management.