Abstract: Taking the Shapotou scenic area as an example, introducing the early warning theory, using the tourism environment bearing rate calculation formula (TECR=TECQ/TECC) and the BP neural network prediction model, the present situation and future of the desert scenic area are analyzed. The conclusion is obtained:(1) The tourism environment bearing capacity early warning border area in the desert scenic area is divided into weak load (<0.8), appropriate load (0.8-1.0), mild overloading(1.0-1.2), moderate overload (1.2-1.5) and serious overloading (>1.5). The corresponding warning states were warned by black light, green light, yellow lamp, orange lamp and red light, respectively. (2) The current tourism environment bearing capacity of the scenic area was in weak state on the whole tourist season, but the May Day Holiday and National Day Holiday are in overloading state. (3) The tourism environmental bearing capacity of the scenic area in the next 5 years will still in the rising trend, but the speed will slow down, the holiday peak period of the May Day Holiday and National Day Holiday will overload. According to the bearing capacity early warning results, the early warning scheme of different.

Key words: tourism environmental bearing capacity, early warning, Shapotou scenic area, desert scenic area