GC53G-1283
How much does downscaling contribute to uncertainty in downscaled climate projections? A case study in Eastern North Carolina.

Friday, 18 December 2015
Poster Hall (Moscone South)
Adrienne Wootten1, Brian J Reich1, Adam J Terando2 and Ryan Boyles1, (1)North Carolina State University Raleigh, Raleigh, NC, United States, (2)USGS Headquarters, Reston, VA, United States
Abstract:
Downscaling is defined as translating the state of variables at a coarse resolution and larger region to the state of variables at a fine resolution for a smaller region. Numerous downscaling techniques exist which are commonly classified as statistical downscaling (based on empirical relationships) and dynamic downscaling (regional climate modeling) techniques. It is common practice for impact assessment and decision making to consider downscaled projections created from one downscaling technique. Several recent studies have demonstrated that different downscaling techniques also have an influence on future projections. In addition, these studies have also demonstrated that the choice of downscaling technique can influence the interim modeling (streamflow, species distribution, etc.) done as part of an impact assessment. Prior studies have also highlighted the three main sources of uncertainty in global climate model (GCM) projections (natural variability, model uncertainty, and scenario uncertainty). However, given the use of downscaled projections and the subsequent demonstration of the uncertainty associated with downscaling, there is a need to consider how downscaling contributes to the uncertainty of future projections. This study presents a methodology to assess the contribution of the four sources of uncertainty in downscaled climate projections (natural variability, GCM model uncertainty, downscaling uncertainty, and scenario uncertainty) using Eastern North Carolina as a case study. This considers future projections of temperature and precipitation across the study domain from six different downscaled projections. It is hypothesized that the contribution from downscaling to the uncertainty of the projections is small compared to the sources from the GCMs which drive downscaling. Results of this study also consider the potential influence on an impact assessment using downscaled climate projections in a region.