Modeling Field of View Effect on the Field Reflectance Measurements for Row Crops

Satellite observations use very narrow field of view (FOV) (less than 0.01deg) but the field measurements use generally very wide FOV (often between 10deg and 40deg) in order to obtain representative sampling size. This difference may introduce some large errors. The objective of this paper is to propose a practical modeling method and evaluate field of view effect on the field reflectance measurements for row crops. The model considers a row crop as a repetition of rectangular walls and then translates to a grid image with sufficient spatial resolution. The wide FOV reflectance is determined by averaging the reflectance of the elements. This model has been used for studying a typical row structure crop (maize canopy) for different observation heights (from 1 m to 5m), different observation angles (from -60deg to + 60deg with step of 5deg) and for three FOV (10deg,25deg,45deg). The results show that wide FOV measurements generally underestimate the reflectance in the red domain (up to -25% of relative reflectance when a 25deg of FOV) and have overestimation in the near infrared domain (up to 10% of relative reflectance when a 25deg of FOV). For different viewing angles, the vegetation contribution is overestimated in the illumination direction and underestimated in the opposite direction; This study can be used not only to analyze that FOV effects, but also to make optimal design of observation geometry (FOV, measurement height, measurement spatial size etc.) for minimizing the measurement errors, and/or to introduce some corrections to reduce the FOV effects.