摘要
以新疆北部牧区为研究区,结合气象台站记录的雪情数据和土地利用类型,对比分析了2001年11月1日-2005年3月31日的MODIS每日积雪产品MOD10A1积雪制图精度。研究表明,1)晴天时MOD10A1产品的精度很高,总精度可达到98.5%,积雪分类精度为98.2%。2)地面台站的积雪分类精度和总精度同海拔之间的相关系数仅为0.04和0.02,但积雪深度与积雪分类精度之间存在显著的相关性。当雪深1~3cm时积雪分类精度为54.1%~94.3%;当雪深3~36cm时,积雪分类精度均大于90%;当雪深大于36cm时,漏测次数为零,积雪分类精度将保持在100%。3)土地利用类型对积雪分类精度有一定的影响。在农田、草原和城市建筑用地3种类型上的总精度分别为97.9%,98.9%和96.9%,积雪分类精度分别为98.0%,98.5%和94.4%。4)3种土地利用类型在不同雪深下的总精度和积雪分类精度都较高。农田、草原和城市建筑用地上的最低积雪分类精度分别为94.6%,95.3%和89.5%,且最低积雪分类精度都出现在雪深为1~10cm的分段上,这个结果与积雪分类精度随雪深的增加而增加相一致。
By the use of NASA EOS Terra/MODIS snow products of MOD10A1, land use type and climatic data, the snow classification accuracy was analyzed for daily snow composite products of MOD10A1 from November 1, 2001 to March 31, 2005 in the northern Xinjiang. Results suggested that:1) Under the clear sky view, overall accuracy of MODIS snow cover mapping algorithm is high at 98.5%, snow accuracy reaches 98.20/40. 2) The correlation coefficient between snow accuracy and elevation is 0.04, and it is 0.02 between overall accuracy and elevation. However, it has a strong correlation between snow accuracy and snow depth. When snow depth is between 1 and 3 cm, the snow accuracy is between 54.1% and 94.3 % ; And when the snow depth is between 3 and 36 cm, the snow accuracy is over 90% ; When the snow depth more than 36 cm, the omission errors is 0 and the snow accuracy keeps at 100%. 3) To some extent, land-use type has influence in snow accuracy. The overall accuracies of cropland, grassland and urban built-up area are 97.9%, 98.9% and 96.9%, the snow accuracies of the three land use types reach 98.0%, 98.5% and 94.4%, respectively. 4) The overall accuracy and snow accuracy of the three land use types under different snow depths are higher. The minimum snow accuracy for cropland, grassland and urban built-up area is 94.6%, 95.3% and 89.5%, respectively. And the minimum snow accuracy for all of land use types appears on 1-10 cm snow depths, which is consistent with the result that snow accuracy increases with snow depth.
出处
《草业学报》
CSCD
2008年第2期110-117,共8页
Acta Prataculturae Sinica
基金
国家科学基金"北疆牧区雪灾监测预警系统"(30571316)资助