A Monte Carlo Assessment of Uncertainties in Heavy Precipitation Frequency Variations

Abstract

A Monte Carlo analysis was used to assess the effects of missing data and limited station density on the uncertainties in the temporal variations of U.S. heavy precipitation event frequencies observed for 1895?2004 using data from the U.S. Cooperative Observer Network (COOP). Based on the actual availability of long-term station data, the effects of limited spatial density were found to be of greater importance than those of missing data. The Monte Carlo simulations indicate that there is a high degree of statistical confidence that the recent elevated frequencies in the United States are the highest in the COOP record since 1895, at least for event definitions using return periods of 5 yr or shorter. There is also high confidence that elevated frequencies seen early in the record are higher than those measured in the 1920s and 1930s, and are not simply an artifact of the limited spatial sampling. The statistically significant shift from high to low values in the early portion of the record, a reflection of natural variability, should not be ignored when interpreting the elevated levels of the most recent decades. Nevertheless, it does appear that the recent elevated levels exceed the variations seen in the earlier part of the record since 1895. The confidence in these statements decreases as the return period increases because of the diminishing number of events in the sample. When a linear trend is fit to the entire 1895?2004 period, the trends are positive and different from zero with a high level of statistical confidence for all return periods from 1 to 20 yr.