Global models are the most widely used tools detecting climate and climate change. However, the model's enormous mathematical complexity limits the geographical resolution that they can provide (1°-5° latitude and longitude) and further limits their capability to capture detailed synoptic and mesoscale weather systems and associated heterogeneous distribution of weather elements. Regional climate models are therefore commonly applied to better represent and understand local weather systems and associated circulations. In this study, the climate simulations over the Arctic marginal seas: Beaufort/Chukchi Seas from a regional model Weather Research and Forecasting Model (WRF) and a global model European Centre for Medium-Range Weather Forecasts (ECMWF) Interim Reanalysis (ERA-Interim) were compared by verifying with the in-situ observations and satellite retrievals for the period of 2005-2009. The preliminary evaluations show that WRF has better capability identifying some local scale weather systems, such as a polar low occurring in the Beaufort Sea in October 2009. In addition, the weather details relevant to topographic effect such as the extreme temperatures over the mountain ranges were also captured well by the WRF model. A detailed evaluation including quantitative statistic analysis between global and regional simulations will be presented at the conference.