Collaborative Research: Quantifying Feedbacks Affecting High Altitude Climate Change

有证据表明，在许多山区，气温变化的速度不同于全球平均水平。由此产生了三个问题：山区气温的上升速度是否快于全球平均水平？在山区，变暖速率是否取决于海拔高度？如果以上的答案都是肯定的，为什么会出现这样的差异呢？几种不同的反馈可以起作用，包括与雪反照率、大气水蒸气、云量和云特性有关的反馈。这些反馈很难量化，因为两个气候变量之间的关系也总是与其他变量相互关联。此外，高海拔地区观测的稀疏性加剧了这一困难。该项目将结合地面和卫星观测与气候模式模拟和神经网络分析方案，以(1)量化导致高海拔地区温度反馈的一些主要关系，(2)研究这些关系和反馈在21世纪如何响应大气温室气体的增加而发生变化。重点将放在青藏高原和科罗拉多州西南部的落基山脉。神经网络分析计算气候变量对之间的偏导数（例如，向下的长波辐射和云量），以便确定反馈回路中各个环节的强度。本研究的广泛影响包括：(1)神经网络可以应用于其他地区，使研究人员能够量化气候系统中的重要反馈并分析非线性过程；(2)结合地面观测和卫星观测，将为研究界提供一个新的空间和时间上扩大的观测数据库；(3)更好地了解山区气候变化将通过改善影响高海拔地区水资源、农业、旅游业和生态系统未来的管理实践使公众受益；(4)将支持一名高中教师与调查人员合作，帮助开发和实施有关山区和气候变化的播客；(5)培养1名博士后和本科生；(6)将与科罗拉多州山地研究所合作编写教育材料。

In many mountain regions there is evidence that temperatures are changing at different rates than the global average. Three questions arise: Are temperatures in mountain regions increasing faster than the global average? Within mountain regions are warming rates dependent on elevation? And if the answers to the above are yes, why do such differences occur? Several different feedbacks can contribute, including those related to snow-albedo, atmospheric water vapor, cloud cover, and cloud properties. These feedbacks are difficult to quantify because the relationship between two climate variables is invariably interconnected with other variables as well. Also, the sparsity of observations in high-altitude regions exacerbates this difficulty.This project will combine surface-based and satellite observations with climate model simulations and a neural network analysis scheme to (1) quantify some of the principal relationships that contribute to feedbacks on temperature in high altitude regions, and (2) investigate how these relationships and feedbacks might change through the 21st century in response to increasing atmospheric greenhouse gases. The focus will be on the Tibetan Plateau and the Rocky Mountains in southwestern Colorado. The neural network analysis calculates partial derivatives between pairs of climate variables (e.g., downward longwave radiation and cloud cover) so that the strength of the various links in a feedback loop can be determined. Broader impacts of this work include: (1) The neural network can be applied in other regions and can enable researchers to quantify important feedbacks in the climate system and analyze non-linear processes; (2) By combining surface-based and satellite observations, a new spatially and temporally expanded observational data base will be available to the research community; (3) A better understanding of climate change in mountain regions will benefit the public by improving management practices that affect the future of water resources, agriculture, tourism, and ecosystems in high altitude regions; (4) A high-school teacher will be supported to work with the investigators to help develop and implement podcasts on mountains and climate change; (5) There will be training for a postdoctoral fellow and undergraduates; and (6) Educational materials will be developed in collaboration with the Mountain Studies Institute in Colorado.