Using Advanced Very High Resolution Radiometer imagery to map rice cropping areas in central Texas

Abstract

In an effort to explore the relationship between satellite imagery and rice production the authors have examined Advanced Very High Resolution Radiometer (AVHRR) imagery which has been acquired for three consecutive growing seasons in an area centered on rice farms located south-west of Houston. The task of relating rice growth to processed satellite data has been broken into two primary tasks. The first is the identification of techniques to assist in successfully utilizing this imagery source to determine the location of rice growing areas. This task is approached by evaluating several image processing techniques on season-long time series of the Normalized Difference Vegetation Index (NDVI) in order to differentiate rice from other vegetative types. The techniques investigated include traditional classification routines applied to the standard and transformed (using a principal component transformation) time series, as well as signal processing techniques such as matched filtering. The matched filter shows good promise, and results from its application to AVHRR are reported. Verification is approached by using ground-truth data obtained from sources such as the Lower Colorado River Authority (LCRA). With a rice cropping area identification technique in place, it will now be possible to investigate how data derived from AVHRR imagery, such as leaf area index (LAI), could be utilized to understand indicators of crop yields. Additionally, since the literature suggests a relationship between crop vigor and emission levels of greenhouse gases, the AVHRR data may, in turn, be used to begin to link satellite image data with emission models on regional scales.