The widespread alpine grasslands and alpine meadows on the Qinghai-Tibet Plateau face the risk of desertification under the influence of climate warming and human activities such as overloaded grazing. Moreover, the alpine and arid climate conditions make the plateau grassland ecosystem extremely fragile and difficult to repair after desertification. It is very necessary to evaluate the spatio-temporal dynamics of land desertification on the Plateau. However, the current method of large-scale rapid desertification simulation monitoring for a long time is insufficient. In this paper, a spatial desertification remote sensing monitoring model based on the vegetation condition-surface reflectance-soil moisture (NDVI-Albedo-Wet) three-dimensional data is constructed through comparative experiments, and the Google Earth Engine remote sensing big data platform is used to implement the model through code. Finally, the dynamic assessment of the long-term spatial pattern of land desertification on the Qinghai-Tibet Plateau is realized. The results reveal the overall mitigation trend of land desertification on the Qinghai-Tibet Plateau, but problems still exist. Specifically, it reflected that:(1) the total area of desertification is decreasing year by year, with an average annual reduction of 551300 hectares, but the desertification simulation data of each phase show that a large number of non-desertification land is transferred to mild desertification. (2) The degree of desertification tends to reverse on the whole, but the problem of land desertification in some regions continues to appear, mainly distributed in "one river and two rivers", the southern foot of Tanggula Mountain, the surrounding area of Tsaidam Basin and Qamdo area. It is also sporadically distributed in Sanjiangyuan, western Sichuan and southern Gansu.