Arctic evapotranspiration: A diagnostic synthesis and model assessment
In this study, the investigators will use available data and modeling tools to understand a key component of the water cycle in a changing Arctic terrestrial environment. Changes in the wetness of Arctic land surfaces can have consequences for vegetative stress, susceptibility to wildfires, and terrestrial release of greenhouse gases such as carbon dioxide and methane. Land surface wetness may also affect wildlife, as well as human activities such as subsistence hunting and fishing, overland travel, access to resources, and water availability. Global climate models are unanimous in projecting increases in Arctic precipitation and nearly so in projecting increases of runoff. Projections of net moisture flux, or the difference between precipitation and evapotranspiration, vary widely among models, especially during the warm season. Observational data point to recent decreases in soil moisture for at least some parts of the Arctic, consistent with increasing evapotranspiration in a warming climate, but increases in surface water in other areas. This study will synthesize different types of existing data to optimize model projections of anticipated changes to Arctic surface wetness. In addition to improved understanding of how a key Arctic process is represented in global models, the broader impacts of this project include training for a graduate student and dissemination of results to policy and stakeholder communities through the Scenarios Network for Alaska and Arctic Planning and Alaska Center for Climate Assessment and Policy programs. The investigators will synthesize three types of existing data on Arctic evapotranspiration (ET): in situ measurements from observing sites, remote-sensing-derived estimates of ET, and global climate model output. They will use these diverse data sources to: 1) clarify the drivers of ET variations over time in different vegetative types and 2) evaluate climate model simulations of terrestrial ET and its variability in the Arctic. The field sites will include various tundra and boreal vegetation types, both with and without permafrost, and observations will include incident radiation, humidity, and wind speed to assess the drivers of ET in the data and in climate models. Global climate model output will come primarily from the Coupled Model Intercomparison Project. Satellite remote-sensing-derived estimates will come from MODIS archives and will bridge the scale discrepancy between models and in situ measurements. Validation of model output against in situ and remote-sensing derived estimates will focus on variation in ET over timescales of days to years. Synthesis of the three types of information will assess key processes that climate models must capture to produce credible projections of Arctic terrestrial hydrology. Based on their ability to capture seasonal cycles and ET over various vegetation types, the investigators will identify the most realistic global models, along with their formulation and structural characteristics. Results will help set priorities for model improvement and model selection to enable more robust conclusions about the trajectory of Arctic terrestrial surface wetness.