Detection and Monitoring of Arid Grazing Land Vegetation using ATSR-2 and Geometric Optical Modelling

This study focused on the Badia region of Jordan and examined how low spatial resolution Along Track Scanning Radiometer (ATSR-2) and Advanced Very High Resolution Radiometer (AVHRR) data can be used to detect and monitor changes in grazing land vegetation in semi-arid environments. Comparisons were made between a wide range of different vegetation indices calculated using the red and near-infrared wavebands of both ATSR-2 and AVHRR imagery and field data collected at a series of sampling sites exhibiting a wide range of percentage vegetation covers. For the ATSR-2 data both the nadir and forward sensor view angles were considered. Poor correlations were found between the vegetation indices and field-measured vegetation cover for the Badia region. It was found that highly reflective soils, very sparse vegetation, and the xerophytic nature of vegetation in and regions limited the use of such indices. Furthermore, due to factors such as angle of view, spatial resolution and problems in geo-location, the forward view of ATSR-2 was not found to add any advantages in terms of the ability to detect and monitor the sparsely vegetated surfaces. As an alternative to vegetation indices, a hybrid geometric optical/empirically based model was developed for the area. Using the information given in the red/near-infrared scattergram of a satellite image, the model allowed percentage vegetation cover to be predicted from the remotely sensed data. Strong correlations (r² = 0.91) were found between the model-predicted percentage vegetation cover and that measured at the field sites. The estimates of percentage vegetation cover derived using the model were similar for both the ATSR-2 and the AVHRR imagery, suggesting that the model results are independent of sensor type. These results suggest that the geometric optical/ empirical model can improve the ability to map vegetation resources in and environments, and that the consistency of the results from the two different sensors should enable long term monitoring of the region.