Submitted to: National Oceanic and Atmospheric Administration
Publication Type: Abstract Only
Publication Acceptance Date: 2/15/2008
Publication Date: 3/4/2008
Citation: Schneider, J.M., Garbrecht, J.D. 2008. Thoughts on assessing decadal-scale precipitation variations as surrogate forecasts [abstract]. In: Proceedings of National Oceanic and Atmospheric Administration. Climate Prediction Application Science Workshop, March 4-7, 2008, Chapel Hill, North Carolina. Available: http://www.sercc.com/cpasw_abstracts.htm Interpretive Summary: Abstract only.
Technical Abstract: For much of the United States, in particular regions without significant ENSO impacts or strong decadal trends, seasonal precipitation forecasts have very limited skill, and forecasts for non-climatological conditions are infrequent. The low predictability for seasonal to interannual precipitation variations in these regions is a subject of intense study, but it seems unlikely that a major breakthrough will occur soon. This situation is particularly disappointing for agricultural interests in the Great Plains and Midwest, prompting a search for alternatives. One possibility is probabilistic guidance based on decade-scale variations in precipitation, a purely statistical approach. There are many problems with such a surrogate, including our inability to forecast the switches between wet, dry, or neutral states. However, it seems possible that guidance based on a thoughtful statistical analysis of decade-scale variations, coupled with a decision concerning current state, could provide more skillful “forecasts” than a climatology composed of the entire record. This presentation will examine the related challenges, and detail an approach for assessing the utility of surrogate forecasts based on decade-scale variations in precipitation.