Quantifying the spatio-temporal variability of Storm Arrival Rates for the continental United States

It is known that earth’s climate exhibits decadal to multi-decadal variability, i.e., cyclic pattern of
warming and cooling, which influence the temporal and spatial changes in the frequency, and
magnitude of events. Based on this, this project aims to determine the trends in precipitation
storms arrival rates (SAR), Extreme storm arrival rate (ESAR) and prolonged event arrival rate
(EAR) for the continental United States. These arrival rates are analyzed to understand their
spatial and temporal manifestations. The data used for the analysis is historic precipitation data
obtained from Global Historical Climatology Network (GHCN), a network of 1244 ground
stations that have 115 years (1900 to 2015) of temporal coverage. To study these aspects the
research incorporates a combination of statistical approaches including imputation of missing
values and estimation of the storm arrival rates. Robust Principal Component Analysis (rPCA) is
used to detect the space-time variability of SAR, ESAR and ESAR across the country.
Preliminary results on principal component analysis show that the first component explains 12%
of the overall variability and indicates a decrease in storm arrival rates for regions in the central
religion of the United States. Using the results obtained coupled with known environmental
conditions, this can be used to infer future local and continental trends which can ultimately help
communities prepare for upcoming events.