空间数据插值方法研究

Research on Spatial Data Interpolation

摘要: 空间数据插值算法是目前普遍应用的一种科学算法，其广泛应用于气象、农业及地质勘探等领域。通过空间数据插值，对缺省或非有效空间数据进行估计和推测，在投资较少的情况下得到大量精度能够满足研究及生产需要的数据，可以在很大程度上满足人类的需要，具有很强的现实意义。本文详细介绍了两种常用的空间数据插值算法——反距离权重插值算法(IDW)和普通克里金插值算法(OK)的原理，并且基于对反距离权重插值算法的研究，针对其可优化参数——距离幂指数k，提出了一种改进的反距离权重插值法。通过数值模拟对济南市的空气质量状况进行空间插值分析，可以得出结论：(1) 普通克里金插值算法优于反距离权重插值算法；(2) 在普通克里金插值算法中，指数模型的插值效果较好；(3) 本文提出的改进的反距离权重插值算法较原算法的插值效果更佳。

Abstract: Spatial data interpolation is a kind of scientific algorithm which is widely used in meteorology, agriculture, geological exploration and some other fields. People can get a lot of data which can meet the needs of research and production with a few investments by using the spatial data interpolation algorithm to estimate the attribute value of unknown point. Therefore, spatial data interpolation can meet the needs of human to a great extent, which has great practical significance. In this paper, we will introduce two classical spatial data interpolation algorithm—the Inverse Distance Weighted interpolation algorithm (IDW) and the Ordinary Kriging interpolation algorithm (OK). Furthermore, on the grounds of researching in the Inverse Distance Weighted interpolation algorithm, an improved Inverse Distance Weighted interpolation algorithm is proposed for the distance power exponent k, which is an optimizable parameter. By applying the spatial data interpolation algorithms which are introduced in this paper into air quality analysis of Jinan city, we can derive the following conclusions: (1) the Ordinary Kriging interpolation algorithm is better than the Inverse Distance Weighted interpolation algorithm; (2) the exponential model has the best interpolation effect in the Ordinary Kriging interpolation algorithm; (3) the improved Inverse Distance Weighted interpolation algorithm is better than the traditional Inverse Distance Weighted interpolation algorithm.