Research On Precipitation Method Of Stone Forest Based On Time Series And High-Order Markov Chain

Rainfall plays an important role in agricultural production,which can reflect changes in droughts and floods in the region.The rainfall amount is used to measure how much rain falls in a region.It can be affected by many factors in nature,such as terrain,climate and altitude.Especially in the past two decades,the intensification of human activities has resulted in many environmental and climatic problems.For example,the greenhouse effect will influence the normal atmospheric rainfall and make the prediction of the rainfall amount more difficult.The main purpose of this paper is to model the meteorological data using the ARIMA model and the high-order Markov chain in time series to study the prediction methods of regional rainfall amount:（1）First,the basic situation of time series,model discrimination process and modeling steps were introduced;then the monthly rainfall data of the Stone Forest area from January 2000 to June 2019 were selected as samples to analyze and test the data;finally,the Eviews software was used to establish time series models for different parameters and predict the short-term rainfall in the Stone Forest area;in addition to months with sudden changes,the overall prediction effect was good,but there were some shortcomings.The longer the time,the lower the accuracy.At the same time,the goodness of fit of the models were not high.It is necessary to consider other factors comprehensively to establish more reasonable models.（2）It is difficult to obtain satisfactory results simply using the first-order Markov chain to predict the rainfall amount.Therefore,this paper introduced the high-order Markov chain for the prediction.First,the first-order Markov was extended to the high-order Markov through relevant properties.Then,starting from the value of the rainfall series,the cluster analysis method was used to divide the series states and calculate the transfer probability.At the same time,the first-order Markov test was extended to the high-order Markov.Finally,the state of the next moment can be predicted in combination with the weigh theory.The model can make accurate predictions for the year with anomaly rainfalls with better effects than the time series.It is suitable for long-term predictions.However,during the prediction and analysis of rainfall amount in the Stone Forest area,the prediction result is only an interval value,which is not accurate.It is necessary to explore the method of predicting the specific value of rainfall amount in subsequent studies.The comparison between the predicted values obtained by the two methods and the actual rainfall values proved that the established model has certain precision and applicability,providing two effective methods for the rainfall amount prediction in the Stone Forest area.