由于积雪和云的反射特性,使用光学遥感监测积雪受到天气的严重干扰,对研究区云量的分析表明,无论是MOD10A1还是MYD10A1,云都是影响该产品对研究区积雪进行实时监测的最大影响因素.综合不同去云方法,利用MODIS逐日积雪产品和被动微波数据AMSR-E雪水当量产品,生成了MODIS逐日无云积雪图像,并利用研究区85个地面气象观测台站提供的雪深数据对合成的单日无云积雪产品进行验证.结果表明:当积雪深度〉3cm时,新产品的积雪分类精度达到91.7%,该产品对实时监测青藏高原积雪动态变化具有重要的使用价值.
Because of the similar reflective charac- teristics of snow and cloud, the weather status can affect the snow monitoring by using optical remote sensing seriously. Cloud amount analysis during 2010-2011 snow season shows that the cloud cov- er is the major limitation to monitoring the snow cover using MOD10A1 and MYD10A1. By the use of MODIS daily snow products and AMSR-E Snow Water Equivalent products, several cloud elimina-tion methods were integrated to produce MODIS daily cloudless snow products. To validate the ac- curacy of the new composited snow product, the information of snow depths from 85 climate sta- tions in the study area was used. The snow classi- fication accuracy of the new daily snow products reaches 91.7% when snow depth larger than 3 cm, which can be used for monitoring snow cover dy- namic change in the Tibetan Plateau.