B53G-0637
Approaches and Recommendations for Simulating Extreme Precipitation Years in Multi-site Experiments
Friday, 18 December 2015
Poster Hall (Moscone South)
Alan Knapp1, Scott L Collins2, Jeffrey Dukes3, Michael E Loik4, Richard Phillips5, Osvaldo E Sala6 and Melinda Smith1, (1)Colorado State University, Fort Collins, CO, United States, (2)University of New Mexico Main Campus, Albuquerque, NM, United States, (3)Purdue University, Department of Forestry and Natural Resources, West Lafayette, IN, United States, (4)University of California Santa Cruz, Santa Cruz, CA, United States, (5)Indiana University, Bloomington, IN, United States, (6)Arizona State University, School of Life Sciences, Tempe, AZ, United States
Abstract:
Worldwide, human activities are exposing all ecosystems to increases in atmospheric CO2, N and temperature. Precipitation also is being altered globally, but increases in precipitation variability and extremes are expected to have greater impacts on ecosystem function than changes in means. Determining how and why ecosystems differ in their sensitivity to precipitation extremes (i.e., drought) is key to forecasting future ecosystem structure and function at the global scale. Coordinated multi-site experiments can be invaluable for assessing differential sensitivity of ecosystems (deserts, grasslands, forests, etc.) to precipitation extremes. However, determining treatment levels in these experiments presents unique problems because extremes in precipitation are defined statistically, based on historical context, and thus can differ dramatically among sites. Therefore, while multi-site experiments with fixed treatment levels may be appropriate for assessing ecosystem sensitivity to CO2 or warming, they may provide less mechanistic insight for studying extremes. We propose that for multi-site experiments focused on variability and extremes, the amount of precipitation removed or added to impose precipitation extremes should be site-specific (not fixed across sites) and matched to the historical climate record. Further, because extreme wet and dry years differ from each other in other attributes (event size, number of events, consecutive dry days, etc.) treatments should incorporate realistic alterations in these precipitation attributes as well. We show that for most ecosystem types globally, experimental infrastructure that passively reduces each rainfall event can realistically simulate drought, with the addition of a few large precipitation events realistically simulating extreme wet years. Thus, while treatment levels required to impose extreme precipitation years should vary among ecosystems, alterations in precipitation attributes can be imposed uniformly.
