Impact of Spatial Scale Choice on Informing Decision-Makers on Changing Characteristics of Rainfall in the United States

Previous studies have revealed statistically significant trends on changing characteristics of rainfall in the conterminous United States using long observational records and precipitation climate extremes indices. Analyses on these changes have proved useful in different water management contexts however, for urban infrastructure design planning, the spatial scale often used to aggregate and present the results impede the ability to readily translate their impact into infrastructure design. For instance, while as a region, the Northeast has experienced an considerable increase in the frequency of heavy precipitation events, some parts of the region might have not experienced same degree of increase. In this study, we analyze the frequency and intensity changes of daily rainfall in the last few decades at different spatial scales. These include climate regions, the Bukovsky regions, climate divisions, as well as state and county boundaries. We then compare the trends found at each scale and evaluate how well they represent the trend at each individual rain gauge station. We then conduct experiments by removing stations within the region and verifying that the surface interpolation provides a change approximate to that of the removed stations. The goal of this study is to make a recommendation on most useful scales to communicate changes in rainfall patterns to stakeholders, infrastructure designers, and decision-makers within a given region.