本文乃筆者博士學位研究之部分，該研究旨在利用衛星遙測學門中的各項影像處理分析技術，應用在美國新墨西哥州半乾燥氣候區，進行地形、生態與環境等自然地理要素的製圖的監測研究，務期能對近百年來在研究區累積的重大環境變遷，諸如灌木侵擾牧地，地壤侵蝕，與水土流失等，作有效並系統地探討與分析。

This research aims to determine the effectiveness of satellite remote sensing for mapping the vegetation change in the form of bush invasion into grasslands caused by land degradation at Jornada in southern New Mexico using mixture modelling (MM), Spectral Vegetation Indices (SVIs) (i.e., simple ratio, NDVI, SAVI, SSI, and PD54), and regression analysis. Bush invasion is one of the most important land degradation factors at Jornada. Historical records pointed out that during the last century, perennial grasslands have been replaced by invasive xerophytic shrub communities. Linear mxture modelling has been used to estimate the proportions of green vegetation, dry vegetation, shade, soil, and rock. The results indicate that mixture modelling is a reasonably accurate measure of these materials. In contrast, many of the correlations between SVIs and the various vegetation parameters are not significant or resulted in low R^2 values. Phenological differences among species were used in estimating grass cover and shrub cover. In dry years such as 1996, in June the shrubs are green but grasses are not. Under these conditions there is a reasonably strong relationship (R^2=0.93) between green vegetation cover and green shrub cover. Dry vegetation cover was also correlated with grass cover (R^2=0.68). Thus, it is possible to obtain the shrub to grass ratio map, which is one of the critical parameters related to bush invasion and associated land degradation in the study area.