多源数据融合计算积雪覆盖数据集——以美国西北部试验区为例

Snow Cover Dataset by Multi-source Data Fusion Algorithm:A Case Study in the Northwestern United States

全面准确地认识积雪动态变化,对制定积雪变化应对措施,合理的进行持续变暖下区域水资源管理,加深全球气候变化科学认识等具有重要意义。作者基于MODIS最新版本的NDSI数据、IMS雪冰数据和192个SNOTEL站点地面积雪观测数据,依据美国西北部积雪特征确定了适于该区域的NDSI积雪判识阈值,针对各数据在不同时间段的积雪判识性能制定了不同的融合规则,提出了多源数据融合算法,得到多源数据融合计算积雪覆盖数据集——以美国西北部试验区(2000-2020)为例。验证结果表明,该融合算法得到的数据不仅较源数据精度有所改善、积雪判识综合性能提高。数据集包括:(1)试验区边界数据;(2)试验区2000-2020年每天的积雪覆盖数据(空间分辨率为500 m)。另附雪深验证点数据。数据集存储格式为.tif、.shp、.xlsx和.txt,共由7,688个数据文件组成,数据量为170 GB(压缩为1个文件,421 MB)。

A comprehensive understanding of snow cover is of great significance with respect to the measurement of changes in snow cover,the development of coping strategies,the management of regional water resources under continuous climate warming,and for acquiring a better understanding of climate change at the global level.Based on the latest MODIS NDSI data,IMS snow/ice data and snow measurements at 192 SNOTEL stations,a suitable NDSI threshold for snow recognition based on the snow characteristics occurring in the Northwestern United States was established.Subsequently,various fusion rules based on data performance for different time periods were formulated.Finally,the snow cover dataset by a multi-source data fusion algorithm for the Northwestern United States region(2000-2020)was developed.A validation study indicated that the data fusion could improve the accuracy and snow recognition performance compared with the source data.The dataset included:(1)the boundary data of the test area;(2)the daily snow cover data of the test area from 2000 to 2020(spatial resolution 500 m).In addition,the data for the verification points for snow depth were included.The formats for storage of the data were.tiff,.shp,.xlsx and.txt,and consisted of 7,688 data files with a data size of 170 GB(compressed into one file,421 MB).