干旱区生态环境知识资源中心

This study evaluated the effectiveness of using Thematic Mapper (TM) data for estimating soil organic carbon (SOC) content in the Zarqa Basin, Jordan, a typical semi-arid environment, and under natural surface conditions by testing a variety of statistical modeling techniques. This is essential for implementing carbon crediting programs for ameliorating the effects of global warming. Although none of the developed models was powerful in predicting SOC, a stepwise regression model was selected since it provided the lowest validation root mean square error (RMSE) of 10.4 metric tons per hectare (ton/ha). Using this model, a SOC map for the basin was constructed by applying map algebra. The total SOC content to 0.2 m depth of the basin was calculated to be 9,423,986.4 metric tons with SOC density of 26.3 ton/ha. This study suggested that, in semi-arid environments and using the statistical modeling techniques that were tested, TM-based SOC models cannot be used for implementing carbon crediting programs; however, they can estimate total surface SOC pools in large areas to within a few percent error.