Paleoclimate record development
We use lake sediments to develop new records of past climate. Recently, we have been developing new study areas in southern Colorado. Winter snowpack in the San Juan Mountains feed the Colorado and Rio Grande rivers, and is particularly vulnerable to rising temperatures and earlier melt seasons. Snowpack in the region is also vulnerable to dust accumulation on snow, which catalyzes earlier snowmelt. In this way dust and drought interact, however we have very few observations of this interaction under natural conditions, which limits our capacity to evaluate how well climate models simulate this feedback. We're developing new records from the abundant lakes and moisture-sensitive tree-ring records in southern Colorado to reconstruct how dust and drought interact over the past several thousand years. Using that new understanding we're collaborating with Toby Ault at Cornell University to investigate how and why climate models simulate that interaction. This research is supported by the Bob and Judi Braudy Foundation.
At the forefront of modern paleoclimatology is the concept of integrating mechanistic modeling with reconstruction and synthesis techniques. The goal is to incorporate scientific understanding of the environmental processes that control how climate variability is recorded by the natural archives we use to understand past changes. For lake systems, this research avenue is largely unexplored. We're collaborating with the University of Alaska Fairbanks are working to develop systems models that simulate how climate variability is filtered through glacier-river-lake systems at three glacier-fed lake basins in the Arctic. As part of this project, we've setup a major new field-based observational network at Lake Peters in the Arctic Wildlife Refuge in the Brooks Range, in the northeast Corner of Alaska. This research is funded by the National Science Foundations Arctic System Science program.
An overarching goal of our research is to understand large-scale changes in past climate dynamics. There are two basic approaches to addressing these questions. The first involves synthesizing paleoclimate evidence for broad regions and estimating of past climate changes in both space and time. We're deeply involved in this approach, developing reconstructions for the past 2,000 years in both the Arctic and more recently in continental-scale regions throughout the world through the international Past Global Changes program. A critical, complementary approach to understanding past climate dynamics is to perform and investigate paleoclimate model simulations with state-of-the-art climate models. Our research focuses on data-model intercomparison, where we aim to get at the best of both worlds by combining the global coverage and detailed climate physics of a climate model with the details of what actually happened from the paleoclimate record.
Our work in paleoclimate synthesis projects has revealed the dire state of paleoclimatic data structure and access. The lack of uniform data formats or standards in the community means that paleoclimate dynamicists spend most of their time on data discovery, retrieval, and formatting, rather than than on investigating the scientific questions of interest. Moreover, this extensive effort that is being needlessly repeated by various research groups around the world. Lastly, critical complimentary data and metadata are typically excluded as they further complicate this data organization process. In an effort to improve on this situation, we have developed a research focus in paleoclimate informatics. Over the past couple of years this research theme has matured considerably, along with collaborators at the University of Southern California. We have multiple projects in this theme, including one focused on developing data interoperability and developing utilities to quantify and visualize the effects of geochronologic uncertainty on paleoclimatic research. Among other advances, this led to the recent publication of the first ever data format for paleoclimatology. We're also developing the Linked Earth wiki, developing methods to crowd-source the development of standardized vocabularies and an ontology using semantic wiki technology.