Future Variations in Cool Season Precipitation along the US East Coast

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner
Tuesday, 4 February 2014: 11:45 AM
Room C101 (The Georgia World Congress Center )
Kelly Lombardo, Avery Point, CT; and B. A. Colle

Cool season, coastal cyclones are capable of producing heavy snow, high winds, and inland flooding, resulting in catastrophic damage along the highly populated eastern U.S. coastal region. Future variations in cyclone frequency, intensity and the associated precipitation will qualitatively and quantitatively alter these regional impacts, influencing the safety of local residents as well as local water resources. Therefore, it is important to understand the variations of these high impact events in the future to allow for adequate preparation. Towards this end, the presented work will assess the ability of climate models to represent past regional precipitation and discuss anticipated future variations in precipitation, especially that associated with coastal cyclones.

Coupled Model Intercomparison Project Phase 5 (CMIP5) data was used to evaluate the distribution and frequency of cool season (November-March) precipitation over the eastern U.S. during the past and future. To assess the ability of the CMIP5 data to reproduce the regional precipitation, fourteen CMIP5 models were compared with the Climate Prediction Center (CPC) Unified precipitation (0.5o resolution) and the CPC merged precipitation analysis (2.5o resolution), for the 1979-2005 historical period. Future analysis will include additional comparisons with precipitation from the Global Precipitation Climatology Project (GPCP) as well. For this 26-year period, the CMIP5 mean accurately reproduces the offshore precipitation maximum associated with the western Atlantic storm track, as depicted in the CPC merged analysis, though the CMIP5 mean over predicts the maximum by 50-150 mm. The CMIP5 mean also over predicts the regional precipitation over the northeastern U.S. by ~100 mm (20%), and is unable to capture finer scale features associated with modifications from lakes and terrain. The over prediction may be due to the tendency for the CMIP5 to over predict higher precipitation events, especially those resulting in >10 mm day-1 of regional precipitation. Despite these issues, overall the CMIP5 relatively accurately reproduces the regional precipitation distribution.

Given the climate models ability to relatively accurately represent both regional precipitation and coastal cyclones (not discussed), analysis was performed to assess the trend in the number of eastern U.S. cool season coastal cyclones as well as the associated precipitation, in both a past (1979-2005) and future (2009-2098) time period. Cyclones were identified utilizing a Hodges (1994; 1995) cyclone tracking approach using 6-hrly MSLP data. Cyclones identified within a specified east coast domain were counted, with the corresponding day classified as a cyclone day. Daily CMIP5 precipitation was summed within the east coast domain on each cyclone day, using 11 CMIP5 members. Through both the past and future periods, there is a decreasing trend in the total number of regional cyclones, though precipitation within the domain occurring on these cyclone days rises, most noticeably throughout the future time period. Furthermore, the CMIP5 mean suggests a increasing trend in heavy precipitation events (>25 mm day-1) over the Northeast U.S, with the largest increase in the late 21st century (2069-2098).