小波最近邻抽样回归耦合模型在三江平原年降水预测中的应用

Wave-nearest Neighbor Bootstrap Regressive Coupling Model and Its Application for Forecasting Annual Precipitation in Sanjiang Plain

近些年来,由于水田面积迅速增加,导致三江平原井灌水稻区地下水位普遍下降,"吊泵"、局部超采现象时有发生。为解决上述问题,以853农场为例,对实测年降水资料进行差分和标准化处理,采用小波理论和最近邻抽样技术进行耦合分析,建立了853农场年降水小波最近邻抽样回归耦合预测模型,精度检验和对比分析结果表明,模型有效性和可靠性较高。该模型揭示了区域年降水量的时间变化规律,为853农场乃至整个三江平原充分利用天然降水、合理制定水稻灌溉制度以及地下水资源的恢复和可持续利用提供了科学依据。

The rapid increase of paddy field acreage has led to widely descending of groundwater level in area of well irrigation paddy in Saniiang Plain recent years. More and more ＂hanging pump＂ and partial overpumping have been occurring in area of well irrigation paddy. In order to solve above-mentioned problems, through taking Farm 853 as example and coupled analysis using the wavelet theory and nearest neighbor bootstrap technology, the wavelet-nearest neighbor bootstrap regressive coupling model that can forecast the annual precipitation was built up after the difference and standardization treatment towards the actual observation annual precipitation datum. The results of precision test and comparative analysis showed this model had upper validity and reliability. The model revealed the time variation regularities of annual pre- cipitation in this area. The study can provide scientific gist for fully utilizing natural precipitation, reasonably establishing irrigation system of paddy, groundwater resource recharge and sustainable utilization in Farm 853 so much as the entire Sanjiang Plain.