Sensibilidade às parametrizações físicas do WRF nas previsões dos parâmetros atmosféricos em shetland do sul e deception

Abstract

The Weather Research and Forecasting model (WRF) was executed with high resolution and three nested grids of 12 km, 3 km and 1 km in the South Shetland and particularly in Deception Island s bay, during the period from 5 to 23 February 2011 at Operation Antarctica 29. Data were obtained in situ by Polar Ship Almirante Maximiano from Brazilian Navy, with a frequency of 1 h. The measured variables are pressure, temperature, dew point temperature, wind speed and direction. The model lateral and boundary conditions were data from NCEP global model, with a horizontal resolution of 1o and updated every 6 hours. Two domain sizes have been tested for the innermost grid, one with 127 and the other with 187 points in each horizontal direction. Besides, two physical configurations have been tested, with the main differences being snow microphysics and soil thermal diffusion. Pressure was the variable best represented by the model, as it is the most dependent on large scale meteorological patterns having only small dependence on the microscale. The wind direction and speed at 10 meters from the surface were also accurate, except when there were sudden fluctuations in wind speed and direction influenced by the topography, in which case there has been an underestimation of the speed. Thermodynamic variables have been very poorly represented by the small inner domain, but improved appreciably when it was expanded. The simulated temperature has an average daily cycle less intense than observed and the underestimation increases in warmer temperatures, probably caused by snow packets in the border areas. The dew point temperature daily average cycle followed the observations. Some observed and simulated patterns in circulation are briefly discussed. Finally, a simulation for a 13-day period during which there were in situ measurements at the Deception Island are presented, showing a good performance when compared to the observations.