LSTM and ARIMA Based Modelling for Arid Environment Studies

This paper focuses on the application of LSTM, ARIMA model and Loteka-Volterra model to plant communities. Firstly, grassland community and soil moisture data were collected to predict the changes of plant communities over time under irregular weather cycles using a gray prediction model, which was implemented by code in MATLAB to calculate the relevant results; secondly, LSTM and ARIMA models were used to study the dynamic changes of plant communities under drought conditions. Then the Loteka-Volterra model was established to explore the changes of soil moisture during plant community turnover. Subsequently, we modified the established Loteka-Volterra model to restore arid environments and change the environmental capacity of non-drought-tolerant plants, after which we established a Loteka-Volterra model that considered the degree of aridity and adjusted the competition coefficient to improve the stability and accuracy of the model. Finally, we considered the effects of pollution and habitat loss and simulated the effects of environmental degradation by competition coefficients in a Loteka-Volterra model that took into account the degree of aridity, thus reducing the environmental carrying capacity.