基于两极冰芯的过去2000年火山活动序列重建研究

Prediction of future climate change depends on the forcing and climate sensitivity of the Earth system. Volcanic eruptions is one of the most important natural causes of climate change, and are among the most critical parameters for determining the climate sensitivity. Despite its importance, existing volcanic indices generally lack the spatial-temporal coverage and resolutions needed in global climate model simulations. Ice cores preserve the volcanic fallouts therefore provide an essential tool for reconstructing past volcanism. But its use depends on solving the two critical questions: how to best integrate the results from multiple ice core records and how to accurately interpolate the spatial-temporal distribution of volcanic radiative forcing that is suitable for climate models. ..The goals of this proposed research project is to produce a volcanic forcing index as a function of month, latitude and altitude for the entire past 2000 years, by combining the 66 ice core records from both Greenland and Antarctic with a new understanding of stratospheric transport of volcanic aerosols. We will apply a multi-state space model to the ice-core-matrix to extract the volcanic signals and calculate the corresponding probabilities. The signals will then be converted to atmospheric sulfate injections by multiplying the coefficients obtained from comparison of the total β activity of the nuclear weapon tests and Pinatubo sulfate deposition in ice cores with their counterparts in the stratosphere. By the time of submitting this proposal, we have collected nine deep ice core sulfate records that extending toward AD 1-500 and beyond. This will be the first time that a volcanic forcing of spatial and temporal dependence to be produced for this early period. The automatic signal extraction and integration procedure will largely reduce the local errors and dating uncertainties associated with individual operations of ice core time series. ..The project will develop an ice core database and spatial-temporal distribution module for volcanic forcing reconstruction. The new forcing index will be submitted to the IGBP PAGES/World Data Center, and distributed freely to the climate modeling community, for use in millennium climate change simulation and attribution studies. The resulting volcanic markers of the past 2000 years, especially those before the 11th century, will provide a valuable tool for ice core chronology, volcano archeology and hazard studies. Results from this project will lead the way to an enhanced understanding of past climate change, which will lead to improve predictions of anthropogenic impacts on climate, clearly of benefit to society.

本项目针对现有火山序列缺乏全球气候模式所需的时间覆盖和空间分辨率的问题，旨在利用66组两极冰芯数据（特别是已收集的9组深冰芯数据），重建覆盖过去2000年且具有较高纬度、海拔和月分辨率的火山强迫序列。多元数据集成及冰芯记录向大气辐射强迫的转化是基于冰芯的火山序列重建中亟待解决的科学问题。拟通过状态空间模型提取信号并计算火山事件的发生概率，解决多元冰芯数据的集成利用问题；通过冰芯中的放射性核试验尘埃和Pinatubo火山沉积量与其相应大气通量的对比，计算从冰芯记录向大气总通量的转换系数；通过经卫星远程观测和模式结果修正的非局部扩散模型，重建具有气候模式所需时空分辨率的火山序列。研究将构建火山活动长序列重建的数据库和时空分布拓展模型，并为全球气候模式提供敏感度计算和气候变化归因分析所需的驱动数据。通过重现过去2000年（尤其是11世纪之前）的火山活动历史，为冰芯年代学和火山考古研究提供依据。

作为气候变化的关键驱动因子，历史火山活动及其辐射强迫序列的重建在区分气候变化的人为与自然因素从而准确预测未来气候变化中起重要作用。本项目针对现有火山序列缺乏全球气候模式所需的时间覆盖和空间分辨率的问题，收集整理了66组两极冰芯数据，修正了全新世晚期的火山活动数据；进一步利用9组长冰芯数据，重建了覆盖过去12000年且具有较高纬度、海拔和月分辨率的火山强迫序列。搭建了状态空间模型提取信号并计算火山事件的发生概率，解决多元冰芯数据的集成利用问题；通过冰芯中的放射性核试验尘埃沉积量以及典型火山事件的模拟与观测的对比，完成了从火山气溶胶的冰芯记录向大气总通量的转换。通过经卫星远程观测和模式结果修正的非局部扩散模型，重建了具有气候模式所需时空分辨率的火山序列，为全球气候模式提供敏感度计算和气候变化归因分析所需的驱动数据。以本项目为第一、二资助，发表SCI论文5篇，培养硕士研究生2人，参加多次国内外学术交流。另外，结合重建和CESM模拟结果，完成了对1815 Tambora和1783Laki等典型火山事件的气候与生态效应研究，提出气候系统内部变率对火山强迫响应的潜在调制作用。在此基础上申请的面上项目获资助。