Leaf Area Index retrieval from multi-annual and multi-polarization SAR time series for crop monitoring

(eng) Reliable, detailed and timely information on crop status and yields is essential for early warning of harvest shortfalls. Satellite remote sensing can provide data that help identify and monitor crops over large agricultural regions. Unlike optical imagery, Synthetic Aperture Radar (SAR) systems present the major advantages of being able to penetrate cloud coverage and being independent of solar radiation. As a result, temporal series of high spatial resolution can be regularly and reliably recorded throughout a growing season. This thesis aims at contributing to the development of a quantitative use of satellite remote sensing in the microwave domain for crop growth monitoring. This work specifically focuses on a methodology development for maize Leaf Area Index (LAI) estimation from SAR data in an operational perspective. LAI is a key variable since it is a state variable in crop growth models that can be coupled with Earth Observation (EO) in the perspective of yield forecasting. Thanks to a very huge data set – including field and satellite data collected over six years, from three SAR sensors, and in two geographical sites – the findings of this research show that the maize LAI retrieval from multi-polarization SAR data is feasible thanks to the semi-empirical Water Cloud Model (WCM). Unlike previously developed methodology, no ground measurements are needed to calibrate the model thanks to the soil moisture estimation from the Soil, Water, Atmosphere and Plant (SWAP) model, and to the synergic use of both optical and SAR images. The calibration of the WCM on anterior growing seasons can also be applied for subsequent seasons. The cross-polarization presents the highest performances to retrieve LAI from multi-polarization data.