The Study Of Stability Of Grassland Vegetation Ecosystem

Grassland desertification have done harm to environment, society and economy, The loss caused by it is increasing. Controlling desertification is a difficult task. Multiple factors such as extreme weather,natural disasters, man-made forms lead to Grassland desertification. That formatted nonlinear ecosystems by means of physical, chemical and biological reaction. The effective tool for the simulation of the grassland ecosystem is the grassland ecosystem dynamics model. Nevertheless establishing a dynamic model must be parameterized. However, model error will exist in the model parameter values obtained via present methods, which may cause the uncertainty of the simulation results of the grassland ecosystem. In this paper, by using Nonlinear Forcing Singular Vector(NFSV) method, Forcing Singular Vector(FSV)method, Conditions of Nonlinear Optimal Parameter-Perturbation(CNOP-P) method, on the basis of the grassland vegetation dynamics model we have studied the model error, and have analyzed the sensitivity of the model when there existed external force. Finally, we analyzed the nonlinear instability of the model.The conclusions are as follows:(1) The grassland vegetation ecosystem of the model that the initial dynamics described has two stable states: desert state and grassland vegetation equilibrium state. As external conditions change, the ecosystem transits between the two states. In this paper, under the condition of joining external forcing, we study the change of dynamic model of grassland vegetation patterns. When joined negative force which gradually decreased, bifurcation interval increased and moved right. With the positive force, when the force increases gradually, the bifurcation interval of bifurcation diagram of precipitation rate R and vegetation quantity B narrows and bifurcation range disappears when f =0.3409. This article only considers when there is a bistable state.(2)Then the influences of two kinds of tendency errors are compared, Forcing Singular Vector(FSV)and Nonlinear Forcing Singular Vector(NFSV), on the basis of grassland vegetation ecosystem model. Due to that different external forces, the bifurcation diagram of the grassland vegetation quantity and precipitation rate is also different. Therefore, this article selected three kinds of typical precipitation rate,and selected different rate of grassland vegetation under the same rainfall rate. For different states,nonlinear changes of development for model error which caused by FSV tendency error and NFSV tendency error are compared respectively. The results show that change of nonlinear developments caused by NFSV tendency error are more obvious than FSV tendency error.(3) Finally we compared the influence caused by NFSV tendency error with CNOP-P method on the model of grassland vegetation dynamics. CNOP-P method represents the influence on mode which caused by the perturbation of model parameter. However, NFSV represent the comprehensive influence caused by various parameters of the mode. Each of the two methods has its advantages.