Research On Glacier Snowline Changes In High Mountain Asia In The Past 30 Years Using GEE

High Mountain Asia(HMA)is the home to the world's most extensive mountain glaciers and snow cover.The rapid glacier retreat and down wasting have a significant impact on regional climate change and water balance.With the continuous and tremendous development of remote sensing,big data,and cloud computing,it has become possible to use remote sensing to research large-scale and long-term glacier changes.Therefore,an automatic algorithm based on GEE,MATLAB,and Arc GIS software is proposed to extract and calculate the glacier snow line altitude(SLA)using multi-source remote sensing data,then compare and evaluate the accuracy of this automatic algorithm with reference glacier.After the evaluation,all glaciers in High Mountain Asia are calculated using this algorithm,and the temporal and spatial variation of snow line altitude in each mountain of HMA are analyzed.The results list as follows.(1)The automatic extraction and calculation algorithm based on GEE,MATLAB,and Arc GIS can accurately calculate the snow line altitude on a large-scale.There are four parts to the algorithm.Algorithm 1 is to remove the cloud impact using GEE.Algorithm 2 is to calculate the glacier accumulation area and ablation area using GEE.Algorithm 3 calculates the snow area ratio(SAR)of the glacier using MATLAB,and algorithm 4 calculates the snow line altitude in Arc GIS and MATLAB.Only several inputs were required to calculate large-scale results before the algorithm calculation automatically.(2)The R² of the observed AAR value of Urumqi Glacier No.1 from 2000 to 2018 and the SAR value calculated by this algorithm is 0.59,which means that the SAR value calculated by the algorithm has a strong correlation with the AAR value obtained by actual observation and calculation.The observed ELA value of the Qiyi Glacier and the calculated SLA value of R² are 0.61,and Urumqi Glacier No.1 Glacier's observed ELA and the calculated R² of SLA are 0.55,which means both of them to have a strong correlation.Meanwhile,the two glaciers'calculated SLA values are closer to the observed ELA values,but the amplitudes of the calculated SLA values are smaller than the observed ELA values.(3)For glaciers in High Mountain Asia,132,541 Landsat images have participated in the calculation,and 42,961 fully qualified glaciers with an area of about 44,979.38 km² are selected.For each mountain in High Mountain Asia,with increasing glaciers'size,the average SAR value range is also increasing,indicating that smaller glaciers are more susceptible to climate change with a greater melting rate.From 1990 to 2020,the overall average SAR of glaciers in the Hindu Kush Mountains is low,and the overall average SAR of glaciers on the Qinghai-Tibet Plateau is relatively high.(4)The increase rate of the glacier SLA anomaly in the Eastern Tianshan Mountains is the largest,2.34±0.32 m a^-1,and the West Kunlun Mountains'SLA anomaly values and Kara-korum Mountains have the smallest increase rates,being 0.33±0.12 m a^-1 and 0.43±0.37 m a^-1,respectively.Among them,the glacier melting rate of the East Tianshan Glacier accelerated after 2005.For the West Kunlun Mountains and Karakoram Mountains,the glaciers are relatively stable,and the glacier SLA anomalies have been negative for many years,indicating that the glaciers were in a positive material balance that year.From a spatial point of view,the glaciers in the Karakoram and West Kunlun Mountains have a lower melting rate,followed by the central and surrounding mountains of the Qinghai-Tibet Plateau,which are mainly at a lower ablation rate.The Tianshan Mountains and the Altai Mountains in the north,the Hindu Kush Mountains,the Alai Mountains,and the Pamirs in the west have a greater ablation rate.In terms of time,the SLA anomalies of many mountain glaciers were mostly negative before 2005 and primarily positive after 2005,indicating that2005 is a possible climate inflection point.(5)Based on the analysis of the interdecadal variation of the glacier SLA of High Mountain Asia,the glacier SLA's spatial distribution presents a zonal pattern of high in the south and low in the north.The highest point of the average SLA is located in the middle of the Qinghai-Tibet Plateau,showing an asymmetric ring.The distribution gradually decreases from the middle to the edge of the Qinghai-Tibet Plateau.Based on the analysis of glacier SLA anomalies from 1990 to 1999,2000 to 2009,and 2010 to 2019,the glacier SLA anomalies from 1990 to 1999 were negative for all-mountain systems,while the glacier SLA anomalies from 2010 to 2019 were positive.The East Tianshan glacier SLA anomaly has the most noticeable change and the fastest growth rate.The glacier anomalies in the West Kunlun Mountains have minor changes.The Karakoram Mountains are the only mountain series with negative glacier SLA anomalies from 2000 to 2009,representing the relatively stable state of the glaciers in the West Kunlun Mountains and the Karakoram Mountains.(6)three main factors affect the SLA calculation results of the algorithm:the accuracy of remote sensing images,the"holes"of SRTM data,and the SLA calculation method.The accuracy of remote sensing images will bring uncertainty to the results.The"holes"of SRTM will affect the SLA calculation results of some glaciers.During SLA calculation,the larger the interval between the elevation bands,the larger the error will be.