Linking X-band radar backscattering and optical reflectance with crop growth models

This thesis describes an investigation into the possibilities of linking X-band radar backscattering and optical remote sensing data with crop growth models for the monitoring of crop growth. The emphasis is on the usability of X-band radar data, with a detailed analysis of the main backscattering influencing factors of agricultural crops in The Netherlands.Six-years of ground-based X-band radar observations (VV and HH polarized, 10° to 80° incidence angle) were used to study the temporal radar backscattering of sugar beet, potato, wheat, barley and oats. The geometry of the crop canopy was found to be a major backscattering influencing factor, especially for the cereals. The possibilities of crop growth parameter (soil cover, biomass, height) estimation from the radar data were investigated using empirical and simple physical relationships. Except for sugar beet in the early growing season, the accuracies of parameter estimation were generally too low to be used in crop growth models.In the optical region, the accuracy of estimating the leaf area index ( LAI ) from vegetation indices was studied. In a case study for sugar beet, the LAI was fairly accurately estimated from the so-called Weighted Difference Vegetation Index ( WDVI ).Two methodologies were developed to link X-band radar and optical remote sensing data with crop growth models. In the first method, remote sensing data were used to estimate the fraction soil cover of a crop as input for a simple lightinterception growth model. This method was especially suitable for the use of optical remote sensing data. The use of X-band radar data was only feasible for sugar beet.In the second method, X-band radar and optical remote sensing data were used to initialize and re-parameterize the crop growth simulation model SUCROS (Simple and Universal Crop Growth Simulator). In six years of sugar beet observations, this method improved the simulation of canopy biomass over the use of SUCROS only. The radar and optical reflectance data were very effective in the initialization of SUCROS, and in adjusting the model during early, exponential crop growth. Optical data also adjusted SUCROS in the middle of the growing season.