Aerial Remote Sensing of Peanut Ground Cover

Abstract

Remote sensing is an effective method for estimating ground cover (GC) of crops. In this study, we compared two methods for estimating the GC of peanut (Arachis hypogea L.). The methods are based on two different vegetation indices, perpendicular vegetation index (PVI) and normalized difference vegetation index (NDVI). We also investigated the use of generalized expressions for the bare soil line and 100% GC point for use in evaluating GC using the PVI method. The field was planted to four varieties of peanuts in two different planting patterns (single and twin rows). High resolution red and near infrared (NIR) images of the study field were collected using the Texas Tech Airborne Multispectral Remote Sensing System (TTAMRSS) on three cloud‐free days during the 2009 growing season. Field measurements of GC were used to compare the two approaches (PVI vs. NDVI). Results showed that GC estimated using the PVI method was in good agreement with corresponding observed values over the range of GC values observed in this study (around 35–95% GC). The NDVI‐based method did not perform as well as the PVI‐based method. An additional finding was that general information involving soil and plant canopy reflectance characteristics obtained from previous studies could be used in the PVI‐based method without significant loss of accuracy. These results support the use of the PVI‐based method for estimating GC of field crops, and show that previously obtained information regarding soil and plant reflectance characteristics can be effectively incorporated into this methodology.