Publications

Ji, QW; Yan, SK; He, Y; Lu, XY; Ma, ZQ (2023). Spatio-temporal patterns of snow cover in the Tien Shan, China from 2000 to 2019 based on cloud-free data supported by Google Earth Engine. REMOTE SENSING LETTERS, 14(3), 265-276.

Abstract
Frequent cloud cover leads to significant problems in estimating snow distribution based on remote-sensing observations, especially in mountainous regions. Supported by Google Earth Engine (GEE), a cloud-gap filling framework using the cubic spline interpolation method was proposed to generate cloud-free data based on the MODIS snow cover dataset. The framework is validated in the Chinese Tien Shan region with in-site snow depth observations, demonstrating good agreement. Subsequently, the spatio-temporal patterns and trends of snow cover in Chinese Tien Shan for nearly 20 years were identified using temporal interpolated snow cover data. The results show that (1) the interpolated images generated by the framework effectively fill the invalid value caused by cloud cover and have significant consistency with the snow depth data from gauges (with the Snow Cover Days (SCD) score of 0.89 and the correlation coefficient (CC) of 0.80); (2) compared with linear interpolation, the interpolated snow cover images generated based on cubic spline method have higher CC with snow depth data from gauges, reflecting higher data reliability; and (3) the snow cover over Chinese Tien Shan demonstrated remarkable periodicity and different spatial distribution, and it had an apparent decreasing trend in the southeastern region of Chinese Tien Shan.

DOI:
10.1080/2150704X.2023.2189030

ISSN:
2150-7058