全球变化下青藏高原积雪动态及其对高寒典型草地植被物候的影响

Plant phenology, as an indicator of environmental variation and climate change impacts, has vital effects on ecosystem biogeochemical cycles. As another major land cover type, snow cover plays an important role in influencing on ecosystem geo-biophysical processes and water availability. For Tibetan Plateau, snow cover and the alpine vegetation are more sensitive to climate change than other regions in China. The seasonal dynamic of snow cover may alter the phenology and affect grassland productivity by modulating soil water availability or other environmental factors. .The pattern and mechanism of snow cover dynamics affecting grassland phenology in Tibetan Plateau is proposed to be investigated, by combining multiple remote sensing index dataset, FLUX observational data and meteorological observational data. The remote sensing phenology propose to achieve by two different vegetation index derived by MODIS. Snow cover duration, snow accumulate period and snow melt period are extracted through moving window detecting method from MODIS snow cover dataset. Linear regressions are used to detect trend of phenology and snow cover season. Partial least square regression and test for Granger Causality are applied to get the time lag effect and causal relationship of snow cover with grassland phenology. Partial correlation coefficient are performed to distinguish whether the responses of phenology to snow are regulate by winter temperature. .The innovations of this proposed project is to reduce the uncertainty of Tibetan Plateau’s grassland phenology retrieve with new remote sensing vegetation index NDPI in order to re-build a precise phenology process from site to regional scales by integrating remote sensing and long-term observations; An another innovation is incorporated into the effects of simultaneously snow’s on plant phenology and consequently separate two distinct snow season over plateau, which could find the time lag effects of snow cover on plant phenology. This project will contribute to enhance the understanding of snow cover effects on ecosystems from an aspect of phenology.

植被物候和积雪特征都是生态系统的关键参数，能直接反映陆地对气候变化的响应，对生态系统碳循环和地-气能量平衡作用重大。青藏高原典型高寒草地和丰富的积雪资源对全球变化敏感，积雪也将影响高寒草地的特征和功能。.以全面分析气候变化下青藏高原积雪动态及其对高寒草地物候的影响为目标，本项目拟利用趋势分析、偏最小二乘回归、格兰杰因果检验和相关分析等方法，以2003-2016年MODIS反演多种植被指数及积雪数据分析青藏高原草地物候受积雪特征驱动的时空变异规律，并采用2011-2016年高寒草甸/草原的地面观测和通量监测数据对变异规律进行验证，拟深入探索积雪对植被物候的驱动作用是否存在时滞效应及该作用是否受温度调节。.研究创新点在于使用NDPI提取植被物候，提高物候反演精度；考虑高原积雪的不连续性；多尺度验证积雪对物候的驱动时滞效应及温度对积雪驱动物候的调控作用，从物候学角度加深积雪对生态系统影响的理解。

植被物候和积雪特征都是生态系统的关键参数，能直接反映陆地对气候变化的响应，对生态系统碳循环和地-气能量平衡作用重大。青藏高原典型高寒草地和丰富的积雪资源对全球变化敏感，积雪也将影响高寒草地的特征和功能。本研究首先分析过去十五年青藏高原积雪覆盖关键特征的时空动态特征，并针对常用植被归一化指数受到生长季前积雪影响导致春季物候提取失真的痛点问题，进而利用多源遥感数据，评估了多种植被指数和荧光遥感产品获取了在高原存在积雪干扰下的植被结构物候和生理物候的能力，筛选出适用于高原存在不连续积雪区域的NDPI指数及CSIF，作为有效提取草地春季物候的主要遥感指标。在此基础上，获取了积雪对春季物候影响的空间影响范围，探究了积雪对春季物候的影响机制，并进一步探究了积雪与春季物候对生长季生产力的驱动作用。主要结果如下： (1) 青藏高原的草地生态系统存在两个积雪高峰时期，第一个为11月底，第二个为2月中旬，67.49%的植被经历积雪覆盖时间缩短的状态，高原中部区域的积雪覆盖时长呈现延长趋势,43.32%的区域存在积雪消融提前；(2) 经通量站点和实地观测验证，NDPI与SIF数据能够有效的排除积雪影响，提取可靠的春季结构物候和生理物候，其中生理物候早于结构物候。PPI可以排除积雪影响提取植被旺盛区的春季物候，但在植被稀疏区观测值过少难以反演物候，而常用的NDVI则受生长季前积雪影响，提取的春季物候相对提前，进而导致反演的高原植被物候提前幅度大于实际； (3) 相比于温度降水环境条件，积雪对春季物候变化起主导作用的区域占高原草地生态系统面积的9.03%，但超过71.15%的区域受积雪变化的驱动，其中相比于积雪覆盖时长，积雪消融时期对春季物候的贡献率更高； (4) 积雪通过调节水分和温度状况对春季物候造成影响，但这种影响作用存在海拔分异和坡向分异特征；在高海拔和南坡向主导地区，积雪改变暖积温进而调节春季物候变异，在低海拔地区和北坡向主导地区，积雪改变冷激温和土壤湿度来调节春季物候变异。同时积雪与春季物候共同作用于土壤湿度对生长季的生产力进行调控。研究结果可以为准确评估气候变化下青藏高原植被动态、碳收支以及畜牧业管理提供科学依据。