局地因素对多年冻土活动层厚度的影响研究—以青藏高原温泉地区为例

As the transition zone of the permafrost and atmosphere, active layer plays a crucial role in water and energy exchange, ecosystem and hydrological processes in permafrost region due to the unique hydro-thermal characteristics of freezing-thawing cycle. Generally, climate regime controls the distribution pattern of active layer thickness (ALT). Local factors, however, such as terrain and vegetation types, may have a greater influence on ALT than that of climate regime under given conditions. Previous studies based on in-situ observations mainly focused on the influence mechanism of individual factor, which insufficiently understand of the correlation among factors and difficultly introduce to ALT distribution model based on these qualitative conclusion. Therefore, the influence of local factors in the regional modelling has not been fully considered, leading to large uncertainties in the mountainous areas. In this project, Wenquan region, located at the eastern Tibetan Plateau, will be selected to explore the influence of local factors on ALT by using field investigation and statistical analysis. The controlling factors will be determined by comparing the different influence of individual factor or factors set on ALT. By consider the synthetically influence of both climate variables and local factors, we develop the distribution model of ALT which is applicable to mountainous areas and complete the simulation and validation on the Wenquan region. The results of this project will be helpful to quantitatively understand the distribution characteristics of ALT in mountainous areas and to provide theoretical basis for accurate regional model simulation.

活动层作为多年冻土与大气间相互作用的场所，其冻融过程中独特的水热特征对地气间能水交换、地表水文和生态过程等产生重要影响。气候背景决定了多年冻土活动层厚度分布的宏观格局，而局地因素在一定条件下其影响可能会超过气候因素。目前基于定位观测的研究主要针对单一局地因子的影响机理和过程，对因子间相关性认识不足，定性分析结果难以量化应用。局地因素的影响在区域模型模拟时尚未得到充分考虑，导致山区的活动层厚度模拟存在较大不确定性。本项目拟在青藏高原东部地形起伏较大的温泉地区，结合野外调查和数理统计方法，开展局地因素对活动层厚度的影响研究。通过对比不同局地因子或因子组合对活动层厚度的影响程度差异，确定主控因子或组合；综合考虑气候和局地因素的影响，建立适用于山区的活动层厚度分布模型，并完成温泉地区区域模拟及验证。研究结果可定量揭示山区的活动层厚度分布特征，为准确的区域模拟提供理论依据。

活动层厚度是表征多年冻土热状态及退化的重要指标，其分布具有强烈的空间差异，尤其在山地多年冻土区，局地因素的影响可能会超过气候背景。确定主控因子并量化不同因子对活动层厚度分布的贡献，对准确的区域模拟和预估具有重要科学意义。本项目通过在青藏高原东部山区布设野外实验，开展了局地因素对活动层厚度的影响研究，通过量化主控因子的贡献建立了山区活动层厚度分布的预测模型。首先，评估了物探方法在研究区的可靠性，实测验证表明探地雷达方法可以准确的获取活动层融化深度和土壤水分，通过建立融化深度与活动层厚度的关系可进一步减小物探方法的系统性偏差。其次，结合实测资料、遥感数据及其它数据产品，制备了研究区内植被类型分布、浅层土壤有机碳分布和融化指数等基础空间数据。通过尺度效应分析确定了土壤水分因子的最佳预测空间分辨率，以植被类型为约束对活动层土壤水分进行了空间模拟。最后，通过量化因子贡献建立了预测模型。结果表明，不同因子组合形式对活动层厚度的解释程度各异，其中包含了土壤水分的因子组合解释率均较高。通过逐步回归去除局地因子间复相关性后，土壤含水量、植被指数、太阳直接辐射量和海拔高度为研究区活动层厚度分布的四个主控因子。利用地理权重回归方法结合四个主控因子建立了活动层厚度分布的预测模型，并在研究区进行了应用。将气候因子加入模型后，对模型精度的提高并不明显。相关研究结果量化了局地因子对活动层厚度的影响，可加深对青藏高原山区活动层厚度空间变化规律的认识，为进一步提高区域模拟精度提供理论依据。