Changes in Extreme Rainfall Event Frequency in the Northeastern United States Using High-Resolution Gridded Datasets

This study relied on several high resolution gridded daily precipitation data sets to analyze the number of extreme precipitation events over the Northeastern U.S. An extreme precipitation event is defined by unique (separated by a minimum of ten kilometers) closed 5.08 centimeter contours as defined by the contour function in the Basemapper module from Matplotlib. The built in get_paths() and vertices functions were used to help define an area enclosed by each contour. This analysis is performed on the PRISM dataset, the satellite derived PERSIANN-CDR dataset and simple nearest neighbor interpolation of station data. Gauge and grid point data are used for further comparison and analysis.

A considerable increase in the occurrence of extreme precipitation at a given point location based on gauge and gridded point data replicated the findings of previous studies. These increases were generally statistically significant, though there was variation in confidence levels between sub regions and time frames. However, when viewed in terms of frequency of extreme precipitation events over the whole Northeastern U.S. (i.e. the annual total number of closed contours), both the high-resolution gridded data and PERSIANN-CDR satellite data both show a steep decline over time. In the PRISM dataset, the inclusion of radar data after 2002 causes an increase in the number of extreme precipitation events detected in the data that must be accounted for. Taken as a whole, the results show that while a given point in the Northeastern U.S. is more likely to see extreme precipitation, the number of extreme precipitation events in the region (as defined by unique closed contours) is sharply decreasing. This implies that the spatial extent of extreme precipitation events may be expanding through time.