地球信息科学学报 ›› 2014, Vol. 16 ›› Issue (3): 470-481.doi: 10.3724/SP.J.1047.2014.00470

• 本期要文(可全文下载) • 上一篇    下一篇

基于TRMM数据的福建省降水时空格局BME插值分析

史婷婷1,2, 杨晓梅1, 张涛1, 刘李3, 田文君4   

  1. 1. 中国科学院地理科学与资源研究所资源与环境信息系统国家重点实验室, 北京100101;
    2. 中国科学院大学, 北京100049;
    3. 中国资源卫星应用中心, 北京100094;
    4. 黄河水利委员会水文局, 郑州450004
  • 收稿日期:2013-09-03 修回日期:2013-10-29 出版日期:2014-05-10 发布日期:2014-05-10
  • 通讯作者: 杨晓梅(1970- ),女,湖北武汉人,博士,研究员,研究方向为遥感与地理信息系统应用研究。E-mail:yangxm@lreis.ac.cn E-mail:yangxm@lreis.ac.cn
  • 作者简介:史婷婷(1986- ),女,河南南阳人,博士生,研究方向为地理空间分析模型研究。E-mail:shitt@lreis.ac.cn
  • 基金资助:

    国家“863计划”课题(2012AA121201、2011AA120101);国家自然科学基金项目(40971224)。

Spatiotemporal Analytical Research of Precipitation in Fujian Province Based on TRMM and BME

SHI Tingting1,2, YANG Xiaomei1, ZHANG Tao1, LIU Li3, TIAN Wenjun4   

  1. 1. State Key Laboratory of Resources and Environmental Information System, Institute of Geographic Sciences and Natural Resources Research, CAS, Beijing 100101, China;
    2. University of Chinese Academy of Sciences, Beijing 100049, China;
    3. China Centre for Resources Satellite Data and Application, Beijing 100094, China;
    4. Hydrology Bureau of Yellow River Conservancy Committee, Zhengzhou 450004, China
  • Received:2013-09-03 Revised:2013-10-29 Online:2014-05-10 Published:2014-05-10

摘要:

传统空间插值方法可获得福建省区域内降水的总体分布,但该地区气象站点较稀疏且分布不均,导致该区域内降水的空间插值结果误差较大。为提高插值精度,本文利用TRMM卫星数据以弥补站点数据的不足,尝试将TRMM数据作为“软数据”、台站数据作为“硬数据”,两者相结合后采用贝叶斯最大熵(Bayesian Maximum Entropy,BME)方法对福建省降水的时空格局进行分析。以2000-2012年近13年20个气象站点的年降水量和月降水量为基础数据,分别利用普通克里格法(Ordinary Kriging,OK)和TRMM为“软数据”的BME插值法,分析福建省多年降水的时空分布格局,并对2种方法的插值结果进行比较。结果表明:在时空分布上,以TRMM数据为辅助变量的贝叶斯最大熵插值结果能更好地体现降水的局部差异特征;在误差评价上,以TRMM数据为辅助变量的贝叶斯最大熵插值结果的MAE和RMSE较小,表明TRMM数据作为“软数据”参与插值的BME方法可以在一定程度上弥补站点数据的不足,有效降低预测结果的绝对误差。通过对福建省降水插值的时空分布格局分析和误差评价可看出,BME插值法通过对基础台站数据,以及TRMM卫星产品数据的利用,使降水的时空分析结果更加真实客观,同时,为TRMM卫星降水数据的应用提供了一个新思路。

关键词: TRMM, 贝叶斯最大熵(BME), 时空分析, 降水, 软数据

Abstract:

Due to the spatiotemporal heterogeneity of regional precipitation, and sparseness and uneven spatial distribution of meteorological stations in Fujian Province, the overall accuracy of the regional precipitation spatiotemporal interpolation is limited. Tropical Rainfall Measurement Mission (TRMM) 3B42 rainfall estimates can directly provide spatially distributed precipitation estimations over global areas, which are used to make up the drawbacks of observed data of rain gauge stations. At the same time, Bayesian Maximum Entropy (BME) is used to improve the accuracy of spatial interpolation because BME can use the soft data with uncertainty and accurate hard data to perform the spatial interpolation. Based on annual and monthly precipitation data of 20 meteorological stations from 2000 to 2012, the BME with TRMM as soft data and the Ordinary Kriging (OK) are employed to analyze the spatiotemporal characteristics of multi-yearly precipitation in Fujian Province, and the results of interpolation methods are compared. The results indicate that BME with TRMM as soft data performs are better in reflecting the local spatial disparity. Comparing the calculation results of two interpolation methods, the MAE and RMSE values of BME with TRMM as soft data are smaller than OK with hard data only. The results show that BME could effectively reduce the absolute error by taking account of using TRMM rainfall estimates as soft data. In conclusion, BME makes the spatiotemporal analysis results of precipitation more objectively because of the integration of hard and soft data. The study also provides new opportunities in application of TRMM rainfall estimates.

Key words: Bayesian Maximum Entropy (BME), TRMM, spatiotemporal, soft data, precipitation