The Qinghai-Tibetan Plateau is sensitive to global climate change and its vegetation is an important indicator of the change. Analyses of the mesoscale spatial variation of vegetation and its spatiotemporal influencing factors in alpine meadows are important for the management and model simulation of degraded meadows. We extracted data for vegetation coverage (VC) from digital photographs of a typically degraded alpine meadow for 22 occasions at 113 locations in a study plot. The spatial and temporal distribution, variability and influencing factors of VC were analyzed using classical statistics, geostatistics, temporal stability and correlation analysis. A first pedotransfer function was built for simulating temporal variation of VC. Average VC was 19.7 and 40.5% in the 2015 and 2016 growing seasons, respectively, and the classical and geostatistical parameters indicated moderate variability and a strong spatial dependence of the VC in the study plot. VC was correlated strongly negatively with soil bulk density and pH and strongly positively with soil organic-carbon density and total nitrogen and total potassium contents. VC was sinusoidally distributed in the 2015 growing season, which was mainly affected by rainfall-induced soil moisture condition, and was parabolically distributed in the 2016 growing season, which was mainly affected by temperature-dominated vegetation phenology. The mean coefficients of variation (CVs) of VC temporal variation in the 2015 and 2016 growing seasons were 39.8 and 36.5%, respectively, which were higher than the CVs (28.0 and 19.1%) for VC spatial variation, indicating that climate affected VC more than soil and topographical properties. The simulation of a pedotransfer function established by temporal variables i.e. 30 cm-depth soil-water content, air temperature and pressure was highly accurate and accounted for >60% of the temporal variability of VC. These results provide basic data and recommendations for the mesoscale simulation of vegetation in alpine meadows and a scientific basis for the management of degraded meadows.

青藏高原对全球气候变化敏感，其植被是变化的重要标志。高寒草甸植被中尺度空间变化及其时空影响因素的分析对退化草甸的管理和模型模拟具有重要意义。我们从一个典型退化的高山草甸的数字照片中，在研究小区的113个位置提取了22次植被覆盖率（VC）的数据。利用经典统计，地统计学，时间稳定性和相关性分析方法，分析了VC的时空分布，变异性及其影响因素。建立了第一个pedotransfer函数来模拟VC的时间变化。2015年和2016年生长季节的平均风投分别为19.7和40.5％，并且经典和地统计学参数表明研究区中的VC具有中等程度的变异性和强烈的空间依赖性。VC与土壤容重和pH呈极显着负相关，与土壤有机碳密度，总氮和总钾含量呈显着正相关。VC在2015年生长期呈正弦分布，主要受降雨诱导的土壤水分状况的影响，在2016年生长期呈抛物线分布，主要受温度主导的植物物候影响。在2015年和2016年生长季中，VC时间变化的平均变异系数（CV）分别为39.8和36.5％，高于VC空间变异的CV（28.0和19.1％），表明气候对VC的影响大于土壤和地形特性。由时间变量（即30 cm深度的土壤水分，空气温度和压力）建立的pedotransfer函数的模拟非常准确，占​​VC时变的60％以上。这些结果为高寒草甸植被的中尺度模拟提供了基础数据和建议，为退化草甸的管理提供了科学依据。