Spatiotemporal Distribution Of Snow And Snowmelt Modeling In Alpine Regions

Snow is an essential factor in cryosphere that is of great importance inChina. The snow shows an indicative response to climate change and affectshydrological cycle system. This study explores the characteristics of the snow atvarious scales: the spatiotemporal distribution of seasonal snow in SouthEastern Tibet and local snowmelt process in a snow column.At the macro-scale, a conceptualized model of Fluctuating Snow-CoveredElevation (FSCE) is proposed, quantifying differentiated onset of snowmeltattributed to different elevation in mountainous regions. The snow in SouthEastern Tibet of approximately15×104km~2is characterized based onmicrowave remote sensing snow product with a spatial resolution of25km. Twokey parameters in the FSCE model, i.e., the median (T_m) and duration (ΔT) ofsnow-free period, are accordingly determined. Meanwhile, the TFPW-MK trendtest is used to reveal the response of seasonal snow to climate change. Theanalysis of27-year time series of T_mand ΔT shows that: the snow-coveredperiod has decreased at low elevations due to the increase in air temperature,and the snow-free period has decreased at high elevations because the increasein precipitation compensates for the increase in air temperature.At the meso-scale, the relationship between snow and elevation and that ofvegetation and elevation are explored for the Yangbajain basin of2665km~2inLhasa River basin. The MODIS snow and vegetation products with a spatialresolution of500m are analyzed. It is shown that there is a pair of "S"-shapedand reversed "S"-shaped profiles for the snow-elevation andvegetation-elevation relationships. This provides an evidence of the closerelationship between the two profiles. It also suggests the effect of spatial scale,that is, the data with the higher spatial resolution provides more information onthe snow-elevation relationship: a non-linear "S"-shaped curve at themeso-scale instead of linear profile at the macro-scale.At the micro-scale, the Snow Column model is developed for representinglocal snowmelt processes in the vertical direction. The observed data at#006 snowpit of Niwot Ridge, Colorado, Front Range of Rocky Mountains, is used tovalidate the Snow Column model. It is shown that the developed model iscapable of characterizing the temporal variation of snow depth, temperature anddensity. Analysis of the response of snowpack to meteorological conditionsshows that there exist two phenomena: one is "the whole snow layer meltsresulting in the decrease in depth and flowing out of water" and the other is"some part of the snow layer melts and the remaining part gets densified". In thefuture, this local scale model could be developed into a spatially distributedmodel through combing spatial snow input and flow routing method.