Estimation of land surface parameters through modeling inversion of earth observation optical data

Abstract

Earth observation (EO) optical data represent one of the main sources of information in the retrieval of land surface parameters (i.e., leaf area index and surface albedo). These parameters are widely used in research and applications in agriculture for improving water and land resources management, especially in the field of precision farming, to monitor crop status, predict crop yield, detect disease and insect infestations, and support the management of farming tasks.During recent years, the technical capabilities of airborne and satellite remote sensing imagery have been improved to include hyperspectral and multiangular observations. In parallel with the advancement of observation techniques, there has been an important development in the study of the interaction of solar radiation with Earth’s surface. This process can be described by using canopy reflectance models of different complexity, which can also be used in operative applications in the field of agricultural water and land management. As such, enhanced EO data and canopy reflectance models can be combined together to reduce the empiricism of traditional methods based on simplified approaches and to increase the estimation accuracy. In this chapter, the application of numerical inversion techniques to a canopy reflectance model is investigated both in the spectral and angular domains. An example of a case study is reported, concerning the estimation of leaf area index in an agricultural site; multidirectional and hyperspectral data, acquired by means of the Compact High Resolution Imager (CHRIS) onboard the Project for On- Board Autonomy (PROBA) platform of the European Space Agency, have been used for the numerical inversion of the canopy reflectance model.