冰川冻土  2007, Vol.29 Issue (5): 722-729  
北疆牧区MODIS积雪产品MOD10A1和MOD10A2的精度分析与评价
兰州大学, 草地农业科技学院, 农业部草地农业生态系统学重点实验室, 甘肃, 兰州, 730020
Accuracy Analysis for MODIS Snow Products of MOD10A1 and MOD10A2 in Northern Xinjiang Area
Key Laboratory of Grassland Agro-ecology System, Ministry of Agriculture, College of Pastoral Agriculture Science and Technology, Lanzhou University, Lanzhou Gansu 730020, China

摘要

以北疆为研究区,结合气象台站记录的雪情数据,利用地理信息系统方法分析了2004年12月1日至2005年2月28日期间北疆地区90个时相的MODIS每日积雪产品MOD10A1和8日合成产品MOD10A2的积雪分类精度.研究表明:1)当积雪深度≤3 cm时,MOD10A1对积雪的识别率非常低,仅为7.5%;积雪深度为4~6 cm时,积雪识别率达到29.3%;积雪深度为15~20 cm,平均积雪识别率达到45.6%.当积雪深度>20 cm时,平均积雪识别率为32.2%;2)MOD10A1产品的积雪分类精度受天气状况的严重影响.在晴空状况下,该产品的最大积雪识别率达到58.2%;但是在多云或阴天时,平均积雪识别率仅为17.8%;3)下垫面对MOD10A1的分类结果也会造成影响,在荒漠区MOD10A1的积雪识别率为39.8%,在草原和稀树草原区的积雪识别率为37.2%,农业用地的积雪识别率最低,为29.1%;4)MOD10A2产品可较好的消除云层对地表积雪分类精度的影响,平均积雪识别率达87.5%,可较好的反映地表积雪的分布状况.

Abstract

By the use of NASA EOS Terra/MODIS snow products of MOD10A1 and MOD10A2 and climatic data,the snow classification accuracy was analyzed using Geographic Information System(GIS) techniques for 90 temporal daily snow composite products of MOD10A1 and 11 temporal eight-day composite products of MOD10A2 from December 1, 2004 to February 28,2005.Results showed that: 1) When snow depth is less than 3 cm,the precision of snow identified by MOD10A1 is very low,only 7.5%;as snow depth is between 4 cm to 6 cm,MOD10A1 snow identification accuracy reaches to 29.3%;and the precision is 45.6% when snow depth is between 15 to 20 cm;The mean accuracy is 31.5% when the snow depth is great than 20 cm;2) the precision of snow identification for MOD10A1 products is severely affected by climatic situation.Under sunshine weather conditions,the snow identification accuracy of MOD10A1 reaches to 50.6%;but the average of snow identification rate was only 18% when it is cloudy or overcast;3) The condition of underlying surface is another factor affecting the MOD10A1 classification results,such as,under clear sky conditions,the precision of snow identified by MOD10A1 is 68% for grasslands with sparse trees and shrubs;in desert,the snow identification rate is 64.4%,and only 40% for agricultural land and 4) It can better eliminate the influence of amount of clouds and improve the snow classification precision for MOD10A2 products,as a result,the mean precision of snow identification is 87.5%,which can reflect better the ground snow distribution and plays an important role in snow disaster monitoring in pastoral areas.
收稿日期:2007-01-03

基金资助

国家自然科学基金项目(30571316)资助

引用本文

[中文]
黄晓东, 张学通, 李霞, 梁天刚. 北疆牧区MODIS积雪产品MOD10A1和MOD10A2的精度分析与评价[J]. 冰川冻土, 2007, 29(5): 722-729.
[英文]
HUANG Xiao-dong, ZHANG Xue-tong, LI Xia, LIANG Tian-gang. Accuracy Analysis for MODIS Snow Products of MOD10A1 and MOD10A2 in Northern Xinjiang Area[J]. JOURNAL OF GLACIOLOGY AND GEOCRYOLOGY, 2007, 29(5): 722-729.
使用本文
PACS
本文作者
阅读笔记
在左边选中内容后,点击→加入笔记。笔记内容将复制到下面文本框中,点击保存按钮可保存在个人文献中心中
              
数据正在加载中...