Collaborative Research: A Late-Glacial Model System for Studying Fine-Scale Vegetational Responses to Abrupt Climate Change

由于温室气体浓度的上升，气候模型预测的未来气候与今天地球上任何地方的气候都不同，换句话说，它们没有现代的模拟物。这样的预测极大地混淆了预测植被反应的努力，因为大多数气候-植被模型依赖于模拟气候。美国中西部的晚冰期，大约在15000到11500年前，也是一个非模拟气候和植被的时期。与今天相比，具有不同寻常的物种关联的植物群落主导了景观。这些“非模拟”群落提供了一个理想的机会，可以在现代封皮之外的条件下研究对快速气候变化的精细响应，并测试旨在预测植被对气候变化响应的模型。该项目将采用多代理方法进行环境重建，包括用于植被重建的化石花粉，用于阐明气候驱动因素的植被建模，以及作为水分平衡独立检查的水文重建。这项研究将建立中西部上游地区晚冰期生态系统变化与北大西洋地区的时间相关性，这是一个长期争论的主题。它解决了古生态学中一个突出的问题，即非模拟化石组合。提出的建模技术旨在预测生物对现代封皮之外的气候的反应，这些气候很可能是温室变暖产生的，对这些气候的预测建模至关重要。

Because of rising concentrations of greenhouse gases, climate models predict future climates that are different from anywhere on Earth today, in other words, they have no modern analog. Such predictions greatly confound efforts to predict vegetation response because most climate-vegetation models rely on analog climates. The late-glacial period in the Midwestern United states, approximately 15,000 to 11,500 years ago, was also a time of non-analog climate and vegetation. Relative to today, plant communities with unusual associations of species dominated the landscape. These "non-analog" communities present an ideal opportunity to investigate fine-scale response to rapid climate change under conditions outside the modern envelope and to test models designed to predict vegetation responses to climate change. This project will employ a multiproxy approach to environmental reconstruction, involving fossil pollen for vegetation reconstruction, vegetation modeling to elucidate climatic drivers, and hydrologic reconstruction as an independent check on moisture balance.This study will establish the temporal correlation of late-glacial ecosystem change in the upper Midwest with the North Atlantic region, a subject of long debate. It tackles one of the outstanding problems in paleoecology, non-analog fossil assemblages. The modeling techniques proposed are designed to predict biotic response to climates outside the modern envelope, climates that greenhouse warming may well produce and for which predictive modeling is crucial.