Abstract:wetlands within alpine inland river basins are highly sensitive to climate fluctuations,readily affected by external disruptions,and inclined to degrade. Timely and precise delineations of changes in alpine wetland areas are vital for enhancing scientific conservation and management efforts. This helps address the sensitivity and vulnerability of alpine wetland resources. The Qinghai Lake Basin,sensitive to global changes,represents a quintessential area of delicate ecosystems on the Qinghai-Tibet Plateau and serves as a distribution zone for internationally significant wetlands. This study leveraged the Google Earth Engine (GEE) platform,employing a long-term dataset of Landsat remote sensing imagery spanning from 2000 to 2023. The random forest classification approach was applied to categorize alpine wetlands within the Qinghai Lake Basin,examining characteristics of their area alterations. Finally,combining correlation analysis and the random forest variable importance ranking method,factors influencing the area changes of alpine wetlands in the Qinghai Lake basin were investigated. The findings disclosed that (1) The mean overall accuracy of wetland classification from 2000 to 2023 was 88.45% (varying between 85.01% and 92.63%),accompanied by an average kappa coefficient of 0.83 (varying between 0.82 and 0.91). In the past,the classification accuracy of swamp meadow was relatively low,with average producer accuracy and average user accuracy at 77.99% and 86.74%,respectively. Lake classification had the highest accuracy,with average producer accuracy at 98.20% and user accuracy at 99.27%. Swamps registered an average producer accuracy of 86.03% and user accuracy of 86.66%,effectively discerning alpine wetland categories including lakes,swamps,and swamp meadows. (2) During the study period,the total wetland area in the Qinghai Lake Basin increased by 604.19 km2,with the swamp areas decreasing by 228.21 km2,while the lake and swamp meadow areas increased by 203.93 km2 and 628.47 km2,respectively. Among the three types of wetlands,swamp meadow showed the greatest variability in the past,exhibiting a fluctuating upward trend,increasing from 21.23% to 28.70% of the total wetland area. Swamp decreased from 6.87% to 2.85% of the total wetland area. The lake area initially decreased between 2003 and 2005,followed by a gradual increase,with a change of 4.65% over time.(3) Driven by increased warmth and precipitation,wetlands exhibited a shift from arid to humid states,primarily marked by transitions between swamps and lakes,as well as between swamp meadows and grasslands. Over 24 years,1101.44 km2 of grasslands converted to swamp meadow,2.93 km2 of lake converted to swamp,and 66.65 km2 of swamp converted to lake. (4) Attributable to variations in climate and precipitation patterns across varying elevations,distinct climatic elements govern the alterations in wetland areas within the Qinghai Lake Basin. In alpine regions,the swamp meadow showed a significant positive correlation with spring snow albedo; the area of lake affected by seasonal snow cover varied. precipitation and seasonal snow cover are the main factors influencing changes in high-altitude wetland areas. The results of this study can provide relevant basis for the scientific conservation and management of alpine wetlands at the basin scale.
