Changing estimates of design precipitation in Northeastern Illinois: Comparison between different sources and sensitivity analysis

Recent research points to changing frequency and intensity of heavy rainfall events and consequently, flood flows in Northeastern Illinois. Flood frequency modeling is a key component of managing floodwaters in this highly populated metropolitan area, which includes Chicago and surrounding urban areas, and has enormous economic and environmental significance. In this research, 100-year 24-h precipitation totals in Northeastern Illinois were quantified using the L-moments method with 1948-2004 hourly precipitation data at six stations in the region for comparison with published results from the US Weather Bureau Technical Paper No. 40 (TP-40), Illinois State Water Survey Bulletin 70 (Bulletin 70) and National Oceanic and Atmospheric Administration Atlas-14 (NOAA-14). Sensitivity analyses were conducted to examine the effects of various factors on 100-year, 24-h precipitation at Aurora College station, in particular the effects of selecting different periods of the precipitation record, different regions, and different underlying distributions. The sensitivity analyses used 1900-2004 daily precipitation data at 12 stations in the region. Finally, the HEC-HMS rainfall-runoff model was used to illustrate relative impacts of changing estimates of design precipitation on flood peaks at 12 small watersheds in the region. It was demonstrated that the oldest source, TP-40, produced significantly smaller 100-year, 24-h rainfall totals, than Bulletin 70, NOAA-14, and the current study. It was also shown that the variability in design rainfall calculated based on different 50-year records (nearly 200%) was much larger than those based on the choice of underlying statistical distribution (50%), or the selection of region (25%). The average relative increase in hydrologic peaks exceeded that of rainfall peaks. This could partly be explained by the non-linear nature of the hydrologic system.