DEM的不确定性对温度插值精度的影响

发布时间：2018-03-11 12:17

本文选题：温度插值　切入点：精度　出处：《地理与地理信息科学》2016年02期 　论文类型：期刊论文

【摘要】：以2001-2010年全国711个气象站点温度观测数据为数据源,在薄板样条插值方法和ANUSPLIN软件支持下,对比分析了使用GTOP30、SRTM3和ASTER GDEM三种DEM数据作为协变量开展空间插值得到的温度成果数据精度。研究发现:1)GDEM的平均相对误差(MRE)、平均绝对误差(MAE)和均方根误差(RMSE)明显高于其他两套数据,数据质量最差;SRTM3和GTOP30的数据精度较高,各误差评估指标大致相同。2)在全国尺度上,在MAE方面,基于ASTER GDEM所得温度插值结果比其他两组DEM得到的温度插值结果高0.4℃左右;在RMSE方面,基于ASTER GDEM获得的温度比其他两组DEM得到的温度高0.5℃左右。基于ASTER GDEM所得温度插值数据的误差明显高于其他两套数据。3)在温度插值误差的空间分布格局上,在我国中东部地区,三种DEM温度插值误差分布规律基本相同;在我国西南部地区,基于GTOP30和SRTM3的温度插值结果的数据精度明显好于基于ASTER GDEM的温度插值结果。4)温度插值误差与DEM高程误差呈现明显的耦合特性,这表明DEM数据的精度是影响温度插值结果的重要因素。
[Abstract]:In the 2001-2010 years from 711 weather stations nationwide temperature observation data, the thin plate spline interpolation method and ANUSPLIN software, analyzed using GTOP30, SRTM3 and ASTER GDEM temperature data of three kinds of DEM data as a covariate in space interpolation precision. The study found that: 1) the average relative error of GDEM the (MRE), the mean absolute error (MAE) and the root mean square error (RMSE) was significantly higher than that of the other two sets of data, data quality is the worst; SRTM3 and GTOP30 data with high precision, the error evaluation index is roughly the same as.2) on the national scale, in terms of MAE, the temperature interpolation than the other two groups of DEM results the ASTER GDEM interpolation based on the results of high temperature 0.4 degrees Celsius; in RMSE, based on ASTER GDEM to obtain the temperature than the other two groups of DEM high temperature of 0.5 DEG. Based on the error of ASTER GDEM the temperature interpolation data significantly Higher than the other two sets of data.3) in the distribution pattern of temperature error of interpolation space, in eastern China, three DEM temperature interpolation error distribution is basically the same; in the southwest region of China, the temperature interpolation results of GTOP30 and SRTM3 based on the data accuracy is significantly better than the temperature interpolation results based on GDEM ASTER.4) the temperature of the interpolation error and DEM elevation error has the obvious characteristics of the coupling shows that, the accuracy of DEM data is an important factor affecting the temperature of the interpolation results.

【作者单位】：中国科学院地理科学与资源研究所;中国科学院大学;中国科学院数字地球与遥感应用研究所;
【基金】：遥感重大专项(00-Y30B14-9001-14/16) 国家重点基础研究发展计划(2010CB9509040102)
【分类号】：P208

【相似文献】