欧亚大陆中部MODIS逐日无云积雪产品（2004-2021秋季）

积雪是全球气候系统的重要组成部分，而云量对遥感积雪产品精度产生了重要影响。本数据集基于MODIS逐日积雪产品，通过MODIS上下午星合成、临近日数据合成、空间相关性合成和MODIS与IMS雪冰合成的多步去云算法结合，制备了欧亚大陆中部（0-160°E，40-80°N）2004—2021年秋季（9—11月）500m逐日积雪覆盖数据集，并利用地面监测站点数据进行了精度验证。结果表明：站点雪深≥2cm时，数据集总体精度达到89.48%，积雪精度达到89.52%，陆地精度为89.47%，高估误差和低估误差分别是9.65%和0.87%，高估误差低于10%的站点有69.75%，低估误差低于5%的站点有85.03%；同时该数据集在森林、草原、耕地农田和城市建设用地的精度较高，灌丛和裸地由于样本较少且积雪覆盖率低导致精度相对低。该研究结果为积雪和气候变化研究了提供科学依据，在预估未来区域气候变化趋势等领域具有重要意义。

Snow cover is a critical component of the global climate system, while cloud cover significantly impacts the accuracy of remote sensing snow cover products. This dataset is developed based on MODIS daily snow cover products, using a combined multi-step cloud removal algorithm that integrates MODIS morning and afternoon satellite composites, adjacent-day data synthesis, spatial correlation-based synthesis, and MODIS-IMS snow and ice composite data. We produced a 500 m daily snow cover dataset for central Eurasia (0–160°E, 40–80°N) during autumn (September–November) from 2004 to 2021, and validated its accuracy using ground-based monitoring station data. The results show that when the snow depth at the stations is ≥2 cm, the overall accuracy of the dataset reaches 89.48%, the snow cover accuracy is 89.52%, and the land accuracy is 89.47%. The overestimation and underestimation errors are 9.65% and 0.87%, respectively; 69.75% of the stations have overestimation errors below 10%, and 85.03% of the stations have underestimation errors below 5%. Meanwhile, the dataset exhibits high accuracy in forests, grasslands, croplands, and urban built-up areas, while showing relatively low accuracy in shrublands and bare lands due to limited samples and low snow cover fraction. The findings of this study provide a scientific basis for snow cover and climate change research, and hold important significance in fields such as projecting future regional climate change trends.

搜集汇总

背景与挑战

背景概述

该数据集是欧亚大陆中部（0-160°E，40-80°N）2004年至2021年秋季（9-11月）的MODIS逐日无云积雪产品，通过多步去云算法生成500米空间分辨率的积雪覆盖数据，具有高精度验证结果（总体精度达89.48%），适用于积雪和气候变化研究。数据以GeoTIFF格式开放获取，大小为5.43 GB，为区域气候趋势预估提供科学依据。