A Methodology for Assessing Extreme Precipitation Trends Applied to Three South Texas Basins, 1898–2011

A statistical scheme based on the Kendall \({\tau}\) statistic, Fisher’s method, the modified Welch t test, and the Gumbel distribution as a special case of the general extreme value distribution was developed to explore the causes of the apparent trends in environmental variables. Daily precipitation data extending as far back as 1898 for gauges in the Nueces, Guadalupe, and San Antonio river basins indicate that the annual maximum daily precipitation depths and the annual 5-day maximum precipitation depth have been generally increasing. The statistical analysis reveals that urbanization, the Pacific Decadal Oscillation, and localized random effects do not contribute to these trends. The results indicate that climate change or other unknown regional phenomena would be the main cause for the increasing trend. The methodology developed can be employed in global change studies to identify the contribution of climate change and other factors, such as societal dynamics, on observed trends in environmental variables.