Trends of extreme precipitation at the North Appalachian Experimental Watershed in east-central Ohio

Abstract

The impact of climate change on many disciplines has been examined by researchers in recent decades. One of the indicators of climate change is increasing extreme precipitation. The annual, seasonal, and monthly precipitation trends at the North Appalachian Experimental Watershed (NAEW) were studied to identify statistically significant changes in precipitation trends and to compare the consistency of local precipitation trends with reported trends on a regional scale. Daily average precipitation was calculated using data collected from 1939 to 2011 at 11 rain gauges located at the NAEW near Coshocton, Ohio. Rank correlation was used to test for trends for three time periods (annual, seasonal, and monthly), for three response variables (total amount of precipitation, number of wet days, and the maximum daily precipitation), and for three data sets (the complete set of data and the upper 10th and 5th percentile data sets). Total precipitation showed a significant increasing trend for the fall season and for the months of October and November. Number of wet days showed significantly increasing trends for the months of August, October, and December. Maximum daily precipitation showed significantly increasing trends for the fall season as well as for three individual months of fall (September, October, and November). Considering the upper 10th, the annual averaged precipitation as well as seasonally averaged precipitation of fall and summer showed significantly increasing trends. Considering the upper 5th, the seasonally averaged precipitation of fall and winter showed significantly increasing trends. Annual and extreme trends on the local scale for the time period of 1958 to 2011 showed consistency with the reported trends by Karl et al. (2009) for regional scales for the time period of 1958 to 2007. The first and second response variables showed significant annual increasing trends for the complete set of data as well as for the upper 10th and 5th percentiles from 1958 to 2011. A piecewise linear least-squares estimation showed that starting in ~1963 annual precipitation changes from a decreasing trend to an increasing trend. However, the decreasing trend was not significant, and only the post-1963 data showed a significant increasing trend. The results were similar in Karl's regional approach. Based on the results, one to three months of the fall season (September, October, and November) have an increasing trend in the three precipitation response variables.