Airborne synthetic aperture radar analysis of rangeland revegetation of a mixed prairie

Abstract

Microwave radar is a potentially useful tool for monitoring the condition of the rangeland. A study was conducted in a mixed prairie community at the Agriculture Canada Research Substation at Onefour, Alberta in 1991 to examine the effects of historical management on synthetic aperture radar (SAR) data obtained from 2 aircraft flights, 24 May 1991 and 1 August 1991. Ground-truthing expeditions were conducted on the same days to obtain estimates of vegetation amounts, species distribution and soil moisture. A former grazing experiment established in 1955 and abandoned 20 years ago enabled comparison of 3 grazing treatments, continuous, rotation and free choice superimposed on native range, crested wheatgrass (Agropyron cristatum (L.) Gaertn.) and Russian wildrye (Elymus junceus Fish.). The ground data and imagery were integrated in a Geographic Resource Analysis Support System (GRASS). Fields that had been cultivated and seeded to Russian wildrye had higher radar backscatter than native range. The radar backscatter from crested wheatgrass fields was similar to native range in May but higher than native range in August. Radar backscatter was positively correlated with number of years since seeding with Russian wildrye. Generally there was little difference in radar backscatter with grazing treatment. Correlation analyses between radar digital number extracted from the ground truth sites and vegetation and soil parameters revealed, depending upon swath mode, significant relationships between radar backscatter and the amount of certain grass species, radar backscatter and canopy moisture, and radar backscatter and soil moisture in May. A significant negative correlation was observed between radar backscatter from the August images, in both swath modes, and percent ground cover. The results of this study indicated a role for SAR imagery in evaluating range characteristics.