基于时间序列分析法的岩溶泉水位预测

济南城区岩溶泉是当地主要的供水水源，查明泉水动态规律并科学合理的预测泉水位对于泉域岩溶水资源的开发利用和保护具有重要意义。本文首先应用时间序列分析法将趵突泉和黑虎泉自2012年5月2日至2018年10月31日的逐日水位数据分解为趋势项、周期项和随机项，分析其水位动态变化规律并建立水位预测模型，结果显示泉水位动态在该阶段无显著趋势性；但受降水的影响，泉水位动态变化呈现两个主要的周期，多年性变化（3.2年）和季节性变化；同时由于受到各种无规律干扰因素的影响，泉水位动态呈现随机波动的随机项。其次，利用2018年11月1日至2020年8月24日的逐日泉水位数据验证上述水位预测模型的预测精度，结果表明该模型运行合理，预测效果较好，具有一定的实用价值。最后利用该模型预测了2020年8月25日至2022年10月31日泉水位动态变化，为当地岩溶水资源开发和管理提供了依据。 

Prediction of karst spring water level based on the time series analysis method

Karst springs in Jinan City are the main water supply source，thus，it is of great significance to find out the dynamic law of spring water and reasonably and scientifically forecast the spring water level for the development，utilization and protection of karst water resources. In this paper，firstly，the daily water level data of Baotu Spring and Heihu Spring from May 2，2012 to October 31，2018 are decomposed into trend terms，periodic terms and random terms by applying with time series analysis method， the dynamic variation law of spring water level was analyzed and the water level forecast model was established. The results show that there is no obvious trend change of spring water level dynamics at this stage，however， under the influence of precipitation，there are two major periods of dynamic variation of spring water level ，perennial change（3.2 years）and seasonal change. At the same time，due to the influence of various irregular interference factors，the dynamic variation of spring water level presents a random term. Secondly，the prediction accuracy of the forecast model is verified by the daily spring water level data from November 1，2018 to August 24，2020，and the results show that the model runs reasonably and has good prediction effect， with certain practical value. Finally，the spring water level（ from August 25，2020 to October 31，2022） is forecasted by above model，which provides a basis for the development and management of local karst water resources.