A long-term 1 km monthly near-surface air temperature dataset over the Tibetan glaciers by fusion of station and satellite observations

[摘要] Surface air temperature (SAT) is a key indicator ofglobal warming and plays an important role in glacier melting. On theTibetan Plateau (TP), there exists a large number of glaciers. However,station SAT observations on these glaciers are extremely scarce, andmoreover the available ones are characterized by short time series, whichsubstantively hinder our deep understanding of glacier dynamics due toclimate changes on the TP. In this study, an ensemble learning model isconstructed and trained to estimate glacial SATs with a spatial resolutionof 1 km  ×  1 km from 2002 to 2020 using monthlyMODIS land surface temperature products and many auxiliary variables, suchas vegetation index, satellite overpass time, and near-surface air pressure.The satellite-estimated glacial SATs are validated against SAT observationsat glacier validation stations. Then, long-term (1961–2020) glacial SATs onthe TP are reconstructed by temporally extending the satellite SAT estimatesthrough a Bayesian linear regression. The long-term glacial SAT estimates arevalidated with root mean squared error, mean bias error, and determinationcoefficient being 1.61  ∘ C, 0.21  ∘ C, and 0.93, respectively.The comparisons are conducted with other satellite SAT estimates andERA5-Land reanalysis data over the validation glaciers, showing that theaccuracy of our satellite glacial SATs and their temporal extensions areboth higher. The preliminary analysis illustrates that the glaciers on theTP as a whole have been undergoing fast warming, but the warming exhibitsa great spatial heterogeneity. Our dataset can contribute to the monitoringof glaciers' warming, analysis of their evolution, etc. on the TP. Thedataset is freely available from the National Tibetan Plateau Data Center at https://doi.org/10.11888/Atmos.tpdc.272550 (Qin, 2022).

[学科分类] 眼科学