Application Of Multi-level Recursive Regression Method In Groundwater Resources Assessment In The Plain Area Of Central Jilin

The plain area of central Jilin province is the most developed area in Jilinprovince, also a base of cash crop in our nation. The fast development ofcivilization and industrialization has brought Jilin Province a miracle in economicdevelopment, but the water shortage has been the uppermost obstacle to centralJilin development. At the same time, with the agriculture modernization and theplantation continue enlargement, the water demand, especially the agriculture waterdemand based on groundwater increase continuously. The groundwater supplyplays a most important role in the central Jilin province, so it is essentiallyimportant to reasonably assess the water resources in this area.The plain area of central Jilin province is spacious but lacking of availablehydrogeology parameter. In the former groundwater assessment, they used onlyone hydrogeology parameter for the whole hydrogeology unit, therefore, theprecision of calculated water resources is always very low, and what's more, theydraw the conclusion that the current groundwater water exploitation quantity is lessthan the permissible exploitation quantity. But from the long term observation dataseries we can see that: the groundwater level in this area has been continuouslydraw down since the mid and later 80's, last century. So, there brings a doubt onwhether the groundwater in this area really under excessive exploitation. This papertries to answer this question by using the stochastic method which avoids using thehydrological parameter to calculate the groundwater permissible exploitationquantity.Since the groundwater exploitation is intense in the study area, thegroundwater system can be assumed as a time-space variation system due to thegreat influence of the exploitation and the precipitation. The Multi-level RecursiveRegression Method is just the one dealing with the time-space variation system. Inthis model, the time-variation parameter can reflect the influence of each input(such as precipitation and artificial exploitation) to the whole groundwater systemand will use concrete data to embody the influence, which shows the time-variationof groundwater system caused by the out world. Because of these advantages, theMulti-level Recursive Regression Method will be applied to modele the regionalgroundwater level in this region. First, predict the time-variation parameter of themodel, and then we can get the different groundwater level under differentexploiting mode, at last we will find the max exploitation quantity that will causethe groundwater level continuously draw down, which is just the permissiblequantity.Our study area located in the central of Jilin province, encompassing a areaof 26,000km2, including nine counties/cities (district): ChangChun city, JiuTai,DeHui, YuShu, NongAn, SiPing, LiShu, GongZhuling, and ShuangLiao. The mainaquifer is the quaternary pore water, which is also our objective aquifer. TheQuaternary deposition is usually less than 20m ,but in the YuShu basin and theinterfluve between east Liaohe and west Liaohe, the quaternary deposition isthicker, with the depth between 20~40m. The sand, sand-gravel, pebble aquifer ismuch thinner, usually less than 10m, and with limited distribution.The mainly research findings in this study as follow:(1) By analyzing the long term observation data in our study area, we find that:the groundwater level in the central Jilin Plain area has been continuously drawdown since the middle and later 80's last century. Draw the groundwater changingmap to figure out the groundwater continuous draw down boundary. Thecontinuous draw down area consist about 90% of the whole study area, amongwhich, the most severe groundwater draw down occurs in the groundwater intenseexploiting area in the city and the paddy field area in the vast agriculture plain.(2) Use the Multi-level Recursive Regression Method to model thegroundwater burying depth changes, and calculate the burying changes underdifferent exploit mode. Through the trending analysis of groundwater buryingdepth, we get the permissible quantity of Quaternary pore water in the central JilinPlain is 1.08 billion m3/n. The detailed distribution is as follow: ChangChundistrict:0.66 billion m3/n, ChangChun city:0.07 billion m3/n, NongAn county:0.12billion m3/n, JiuTai city:0.08 billion m3/n, DeHui city:0.13 billion m3/n,YuShu:0.26 billion m3/n, SiPing district:0.42 billion m3/n, ShuangLiao:0.12 billionm3/n, LiShu:0.18 billion m3/n, GongZhuling:0.12 billion m3/n.(3) All the nine counties/cities exist over pumping problem at different degree.