Remote sensing of water pollution has constituted an important research field in recent years. This

paper intends to analyze water pollution of the Aksu-Tarim River section using TM data. A model has been established

between the DN values of TM images and the ground water pollution data obtained from the local water monitoring

agency along four cross sections of the river. The results indicate that there exists a good correlation between the

DN values and the ground data, suggesting that the model is applicable for the estimation of water quality in the

river. Using the model, the authors tried to re-build the water quality variation curves of the river from the

available TM images. This study provides another example for the argument that remote sensing is a powerful

technique for water pollution monitoring. It is highly possible for remote sensing to assess water quality in a

river system when high-quality spatial and temporal remote sensing data required are available.