基于季节ARIMA模型的地下水位动态预测

Groundwater Regime Forecasting Based on Seasonal ARIMA Model

通过分析1996～2004 年测井实测逐月地下水埋深序列的趋势性和周期性,建立了合适的季节ARIMA 模型,并运用该模型对2005 年逐月地下水埋深进行预测,预测值与实测值变化趋势一致,平均相对误差为2.5%,证明该模型用于地下水水位预报是可行的.

By analyzing the intrinsic trend and periodicity of the situ groundwater level data from 1996 to 2004, an appropriate seasonal ARIMA model was established to forecast the groundwater variations in 2005. After comparing with the actual value, the results were consistent with the situ data, with the average relative errors of about 2.5%. Evidences from the comparison show that the seasonal ARIMA model is suitable for forecasting the groundwater regime.