Extending the Predictability of Rain Shadow Effects Using Spatial Analogs

The goal of this study is to improve forecasts of precipitation and cloudiness over southwest British Columbia, Canada. We investigate the predictability of a Rain Shadow Effect (RSE), characterized by atmospheric states in which the downstream precipitation and associated cloud mass are suppressed as a consequence of flow interaction with orography. When upstream low-level flow is perpendicular to the steep terrain over Vancouver Island, downsloping winds and mountain waves develop over the Strait of Georgia. This condition favors a RSE, characterized by cloudiness and precipitation surrounding the metropolitan Vancouver. The RSE is given by a nonlinear relationship between the atmospheric flow and topography. Consequently, to forecast the RSE, NWP models must correctly predict low-level atmospheric flows, and must accurately represent the region's complex topography. Forecasting the RSE over Vancouver is specifically challenging because of the lack of low-level observations over the eastern Pacific in the so-called Pacific Data Void. The sparse upstream observation network weakens forecast skill normally obtained from data assimilation, and prohibits thorough model verification. Consequently, the atmospheric profile associated with the RSE remains unknown. In this study, analog forecasts are created with a similarity measure based on spatial flow patterns obtained from infrared satellite channels. RSE occurrences are identified using radar reflectivity observations, and the corresponding mesoscale spatial pattern from satellite imagery provide the 2-D flow portrait to search for analogous patterns. This analysis will indicate whether spatial similarities in water vapor and cloudiness flows may be used to detect predictive features and inform probabilities of developing rain shadows over the metropolitan Vancouver region.