Submitted to: Climate
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 2/23/2019
Publication Date: 2/27/2019
Citation: Joshi, S., Garbrecht, J.D., Brown, D.P. 2019. Observed spatiotemporal trends in intense precipitation events across United States: applications for stochastic weather generation. Climate. 7(3): 36. https://doi.org/10.3390/cli7030036.
DOI: https://doi.org/10.3390/cli7030036 Interpretive Summary: The study presents information on definition and categories of intense precipitation (all precipitation events above 90th percentile in a precipitation distribution) and techniques used to derive thresholds of such categories. It also assesses categorical and observational trends of intense precipitation at regional and national scales. Intense precipitation has been widely divided into three categories: heavy, very heavy, and extreme. There are two approaches to determine thresholds for these categories: fixed numerical and percentile-defined. Although national trends in these categories have been reported, the regional trends lack clear distinction. Site specific trends in intense events of estimate precipitation distributions can be inaccurately assessed due to this. Analysis of probability exceedance curves of observed and forecasted distributions can help resolve this issue.
Technical Abstract: An increasing focus of climate change studies is the projection of storm events characterized by heavy, very heavy, extreme, and/or intense precipitation. Projected changes in the spatiotemporal distributions of such intense precipitation events remains uncertain, due in large measure to variability in both the definition and evidence of increased intensity in the upper percentile range of observed daily precipitation distributions, particularly on a regional basis. As a result, projecting changes in future precipitation at the upper tail of the distribution (i.e., the heaviest events), such as through the use of weather generator models, remains challenging. One approach to address this challenge is to better define what constitutes intense precipitation events, and the degree of location-specific adjustment needed for weather generator models to appropriately account for potential increases in precipitation intensity due to climate change. In this study, we synthesized information on categories of intense precipitation events, and assessed reported trends in the categories at national and regional scales. Investigations of adjusting weather generation models to include long term regional trends in intense precipitation events are limited and modeling trends in site specific future precipitation distributions forecasted by weather generator programs remains challenging. Probability exceedance curves and variations between simulated and observed distributions can help in modeling and assessment of trend in future extreme precipitation events that reflect changes in precipitation intensity due to climate change.