Impact of moisture flux and freezing level on simulated orographic precipitation errors over the Pacific Northwest

Two cool seasons (November-March) of daily simulations using the fifth-generation Pennsylvania State University-NCAR Mesoscale Model (MM5) over the Pacific Northwest are used to investigate orographic precipitation bias. Model simulations are compared with data from a radiosonde site at Salem, Oregon, just upstream (west) of the Oregon Cascades; precipitation gauges over a portion of the Pacific Northwest; and a National Weather Service Weather Surveillance Radar-1988 Doppler (WSR-88D) in Portland, Oregon. The 77 storms analyzed are partitioned into warm/cold storms based on the freezing level above/below the Oregon Cascades crest (;1600 m MSL). Although the seasonal precipitation is well simulated, the model has a tendency to overpredict surface precipitation for cold storms. The correlation between the upstream relative humidity-weighted integrated moisture transport and precipitation for warm storms (r 250.81) is higher than that for cold storms (r 2 5 0.54). Comparisons of model ice water content (IWC) and derived reflectivity with radar-retrieved IWC and observed reflectivity for the 38 well-simulated storms show reasonably good agreement for warm storms but an overprediction of IWC and reflectivity aloft for cold storms. One plausible reason for the persistent overprediction of IWC in cold storms might be related to the positive bias in snow depositional growth formulation in the model bulk microphysics parameterization.Afavorable overlap of the maximumsnow depositional growth region with the mountain wave ascent region in cold storms magnifies the bias and likely contributes to the precipitation overprediction. This study also highlights the benefit of using data aloft from an operational radar to complement surface precipitation gauges for model precipitation evaluation over mountainous terrain.