Ground-truth-aided crop monitoring

The development of efficient methods for analysis and synergistic use of multisource data is one of the most essential and challenging issues that the remote sensing community faces recently. The importance of this issue is relevant to the ever-increasing quantity of data and the diversity of targets and investigation goals. Entering wider into their operational stage, remote sensing technologies face higher requirements to the accuracy of the information they provide. Because of the raising need for reliability of the information products, ground-based observations are considered one of the pillars of remote sensing and are used as a reference source for the development and validation of retrieval algorithms. The interpretation of airborne and satellite data provided by numerous sensors and applied in multipurpose environmental studies requires comprehensive knowledge of land covers spectral behavior. Obtaining quantitative information about vegetation is among the application priorities of Earth observations. In agricultural monitoring, ground-truth studies of crop performance and spectral response at different development stages and under different conditions are essential for the improvement in modeling activities. The reliability of the output estimations is of prime importance in implementing the derived information for crop condition evaluation, growth assessment, stress detection, and yield forecast. In view of this, our paper presents an approach for the verification of spectrally-based retrievals of crop growth variables and yield predictions. Ground-based and airborne experiments have been conducted to validate the performance and assess the predictive accuracy of crop spectral models. Results are presented of evaluating the correspondence of crop variables output obtained from spectral and from biophysical relationships. Models using spectral features as predictors of canopy parameters are examined for their predictive applicability. The algorithm has been tested on experimental data over winter wheat fields and has revealed the rationality of the approach.