Comparison of MM5's simulations with other RCMs from the NARCCAP multi-model ensemble simulations in Central Alberta

Abstract

A regional climate model, MM5, is used to simulate summer precipitation of central Alberta. The model is forced with ERA-Interim reanalysis data and run using a one-way two-domain nested framework, with an outer domain resolution of 9 km and an inner domain resolution of 3 km. Annual Maxima of precipitation (AMP) data are required to develop Intensity Duration Frequency (IDF) curves necessary for designing municipal hydrologic infrastructures. The AMP values of 3-, 6-, 12-, and 24-hr durations for 2-, 5-, 10-, and 25-yr return periods, are computed from MM5's high resolution output. First, simulated rainfall statistics were compared to those of local rain gauge data obtained from the City of Edmonton. The analyzed results indicate that short-duration (15-min to 1-hr) quantiles of AMP calculated from MM5 simulated rainfall, are similar to observation, while longer-duration (2-hr to 24-hr) quantiles of AMP are overestimated. The simulated average rainfall intensity is generally representative of observation but the total overall rainfall amount and frequency has a large positive bias. Second, to evaluate model uncertainties, the MM5 rainfall data were compared to those of other regional climate models (RCMs) from the North American Regional Climate Change Assessment Program (NARCCAP), which were driven by NCEP reanalysis data. Our results show that MM5's performance, in terms of average rainfall intensity and total rainfall frequency, is comparable to that of other RCMs and that they are within the bounds of intra-model uncertainty. However, the quantiles of AMP and total rainfall amounts are outside the bounds of intra-model uncertainty.