测绘学报 ›› 2016, Vol. 45 ›› Issue (4): 458-465.doi: 10.11947/j.AGCS.2016.20150123

• 地图学与地理信息 • 上一篇    下一篇

顾及时空异质性的缺失数据时空插值方法

樊子德1, 龚健雅2, 刘博3, 李佳霖1, 邓敏1   

  1. 1. 中南大学地球科学与信息物理学院, 湖南长沙 410083;
    2. 武汉大学测绘遥感信息工程国家重点实验室, 湖北武汉 430079;
    3. 日电(NEC)中国研究院, 北京 100084
  • 收稿日期:2015-03-09 修回日期:2016-02-02 出版日期:2016-04-20 发布日期:2016-04-28
  • 通讯作者: 邓敏,E-mail:dengmin208@tom.com E-mail:dengmin208@tom.com
  • 作者简介:樊子德(1988-),男,博士生,研究方向为时空数据插值与建模。
  • 基金资助:
    国家863计划(2013AA122301);湖南省博士生优秀学位论文(CX2014B050);中南大学研究生创新项目(2015zzts067)

A Space-time Interpolation Method of Missing Data Based on Spatio-temporal Heterogeneity

FAN Zide1, GONG Jianya2, LIU Bo3, LI Jialin1, DENG Min1   

  1. 1. School of Geosciences and Info-Physics, Central South University, Changsha 410083, China;
    2. State Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University, Wuhan 430079, China;
    3. NEC Labs, Beijing 100084, ChinaAbstract
  • Received:2015-03-09 Revised:2016-02-02 Online:2016-04-20 Published:2016-04-28
  • Supported by:
    The National High Technology Research and Development Program of China(863 Program)(No.2013AA122301);The Hunan Funds for Excellent Doctoral Dissertation(No.CX2014B050);The Central South University Funds for Excellent Doctoral Dissertation(No.2015zzts067)

摘要: 时空插值方法被广泛应用于缺失时空数据集的插值与估计。时空插值是时空建模与分析的一个重要内容,当前该研究关注的热点之一是异质条件下的时空插值与估计问题。因此,本文从时空数据的异质性出发,提出了一种顾及时空异质性的缺失数据时空插值方法。该方法首先对数据集进行时空分区,然后分别在时间和空间按照异质协方差模型计算缺失数据的估计值,进而利用相关系数确定时空权重、融合时间和空间估计值得到缺失数据的最终估计结果。最后通过两组气象数据集进行交叉验证对比分析试验。试验结果表明本文方法对比其他插值方法具有更高的精度和适用性。

关键词: 时空插值, 分区, 异质性, 缺失数据

Abstract: Space-time interpolation is widely used to estimate missing data in a dataset integrating both spatial and temporal records. Although space-time interpolation plays a key role in space-time modeling, it is still challenging to model heterogeneity of space-time data in the interpolation model.To overcome this limitation, in this study, a novel space-time interpolation method based on spatio-temporal heterogeneity is proposed to estimate missing data of space-time datasets. Firstly, space partitioning and time slicing of space-time data was implemented. Then the estimates of missing data are computed using space-time surrounding records with heterogeneous spatio-temporal covariance model.Further the weights of space and time are determined using the correlation coefficient and the finally estimates of missing data is combined integrating time and space estimates. Finally, two datasets are selected to verify the accuracy of this method. Experimental results show that the proposed method outperforms the four state-of-the-art methods with higher accuracy and applicability.

Key words: spatio-temporal interpolation, partitioning, heterogeneity, missing data

中图分类号: