基于ARIMA预测的草原放牧策略——以锡林郭勒草原数据为例

Grassland Grazing Strategy Based on ARIMA Prediction—Taking the Data of Xilingol Grassland as an Example

摘要: 自“退牧还草”政策以来，我国在保护和改善草原生态环境方面取得了显著成效。放牧作为草原规划的重要部分，其合理规划有助于草原生态环境的稳定，实现草原的可持续发展。本文以锡林郭勒草原2012~2022年的土壤与降水数据为例开展研究。首先，以草原100 cm深度的土壤为对象，建立放牧强度与土壤湿度的放牧策略模型。其次，基于放牧策略不变的前提下，结合历年土壤与降水数据和土壤含水量与牧区供水率和土壤植被覆盖率之间的动力学关系模型，利用ARIMA时间序列预测法对锡林郭勒草原未来土壤与降水进行预测，建立土壤湿度预测模型，并预测出土壤在当前放牧策略下未来20个月内草原土壤100 cm深度的湿度水平。ADF检验和KPSS检验表明，预测结果的置信度水平为95%，可以有效刻画放牧强度与草原地下水资源的关系。本文的研究结果可为草原规划中的放牧策略选择提供参考。

Abstract: Since the policy of “returning pasture to grassland”, China has achieved remarkable results in protecting and improving the grassland ecological environment. As an important part of grassland planning, reasonable planning of grazing helps to stabilize the grassland ecological environment and achieve sustainable development of grassland. This paper takes the soil and precipitation data of Xilingol grassland from 2012 to 2022 as an example to carry out research. Firstly, taking the soil at 100 cm depth of the grassland as the object, the grazing strategy model of grazing intensity and soil moisture is established. Secondly, based on the premise that the grazing strategy remains unchanged, combined with the soil and precipitation data of previous years and the kinetic relationship model between soil water content and pasture water supply rate and soil vegetation cover, the future soil and precipitation in the Xilingol Grassland were predicted using the ARIMA time series prediction method, the soil moisture prediction model was established, and the soil moisture level was predicted for the grassland soil in the next 20 months under the current grazing strategy. The ADF test and KPSS test show that the confidence level of the prediction results is 95%, which can effectively portray the relationship between grazing intensity and grassland groundwater resources. The results of this paper can provide a reference for the selection of grazing strategies in grassland planning.