Desertification is one of the most alarming processes of environmental degradation. At present, people always combine two methods together to study the desertification. The), evaluate the desertification degrees in some small-area discrete points with the use of Ground-based Monitoring,, and then analyze the areas in RS image whose desertification degrees are similar to the discrete points. Therefore, the desertification degree of the whole target area is obtained ultimately This is feasible but it needs many different soil samples combined with RS image. At the same time. lots of manual intervention is still needed. In this paper, we design a new automatic system called TurningGreen which can monitor the desert areas. evaluate the desertification degree and forecast the tendency, of desertification in a small scope of area. In this system. Wireless Sensor Node (WSN) is used to collect the original meteorological information, which is used to evaluate the desertification degree and forecast the desertification tendency in the target area. The wireless sensor nodes communicate with each other using ZigBec protocol, and finally form a dynamic route tree with sink (a wireless senor node with an additional serial port) being the root node. The eBoxII is connected to the WSN sink, it is used as the gateway of' the system. and runs sonic core algorithm such as Map Generation, the ANN and Markov Chains in order to accomplish the Map Generation, Desert ificat i on Status Display and Desertification Tendency Display of the target area.