SIMULATION OF THE MODERN ARCTIC CLIMATE BY THE NCAR-CCM1

The NCAR CCM1's simulation of the modem arctic climate is evaluated by comparing a five-year seasonal cycle simulation with the ECMWF global analyses. The sea level pressure (SLP), storm tracks, vertical cross section of height, 500-hPa height, total energy budget, and moisture b udget are analyzed to investigate the biases in the simulated arctic c limate. The results show that the model simulates anomalously low SLP, too much storm activity, and anomalously strong baroclinicity to the west of Greenland and vice versa to the east of Greenland. This bias i s mainly attributed to the model's topographic representation of Green land. First, the broadened Greenland topography in the model distorts the path of cyclone waves over the North Atlantic Ocean. Second, the m odel oversimulates the ridge over Greenland, which intensifies its blo cking effect and steers the cyclone waves clockwise around it and henc e produces an artificial ''circum-Greenland'' trough. These biases are significantly alleviated when the horizontal resolution increases to T42. Over the Arctic basin, the model simulates large amounts of low-l evel (stratus) clouds in winter and almost no stratus in summer, which is opposite to the observations. This bias is mainly due to the locat ion of the simulated SLP features and the negative anomaly of storm ac tivity, which prevent the transport of moisture into this region durin g summer but favor this transport in winter. The moisture budget analy sis shows that the model's net annual precipitation ([P - E]) between 70-degrees-N and the North Pole is 6.6 times larger than the observati ons and the model transports six times more moisture into this region. The bias in the advection term is attributed to the positive moisture fixer scheme and the distorted flow pattern. However, the excessive m oisture transport into the Arctic basin does not solely result from th e advection term. The contribution by the moisture fixer is as large a s from advection. By contrast, the semi-Lagrangian transport scheme us ed in the CCM2 significantly improves the moisture simulation for this region; however, globally the error is as serious as for the positive moisture fixer scheme. Finally, because the model has such serious pr oblems in simulating the present arctic climate, its simulations of pa st and future climate change for this region are questionable.