Simulation of snow storm of 6-12 February 2005 in Tehran with Advance Regional Prediction System (ARPS)


During February of 2005 the north of Iran witnessed the development and activity of a variety of severe meteorological systems accompanied with unusual heavy precipitation, mostly snow and hail. The rate and amount of snow in the northern coastal region was so heavy that it made a new record for that region with a report of more than 3m of snow in Rasht, a city located in that area. Also the damage it caused in the area was great, with heavy human and economic losses. In this region with a dominant subtropical and sub-Saharan climate such an event is very rare and the yearly average of precipitation is much less than the global average, therefore, occurrence of such systems should be treated as a very motivating case for meteorologist and atmospheric researchers. The rate of precipitation is not the only factor that makes it a particular case but also the duration (more than 2 weeks) and type of precipitation (frequent snow accompanied with hail-storms) was unusual. The domain of activity of these systems is also a challenge, covering one third of the north of Iran and the east of Turkey and the west of Afghanistan, a domain in the Middle East where the prediction of atmospheric events is highly risky because of its topological existence factors influencing weather and precipitation. This system was activating in the area from Feb. 5-12 and because of its intense precipitation was chosen as the case to be simulated.
Our purpose in carrying out this study is to explore through sensitivity experiments, the possible success of the applied numerical forecasting model. The results of these experiments are of interest for what they reveal both the specific storm and the predictability of precipitation within storms. Predictability of large scale meteorological parameters, especially the precipitation, is limited to within 2 weeks, but for mesoscale (where precipitation is one of the most interesting and most challenging parameter for a model) prediction, it is not clear how long and how successful it will be. These questions might be asked from 2 points of view, first the previous research and simulations of systems in this area (the Middle East) and second, the ability and quality of the numerical model for simulation of such systems. In this study the Advance Regional Prediction System (ARPS) model is run with horizontal resolution of 4, 10 and 30km in three phases for 6-12th Feb. 2005 and the results have been analyzed. In 30km horizontal resolution running at the surface level and 500 hPa levels have been compared with observation. Generally, in sea surface maps, the pressure simulated by the model is less than the actual one. But there is a reasonable similarity between the 2 map patterns. The simulated geopotential height in 500hPa level is in agreement with observations. For the 10km simulation, the isoprecipitation maps of the northern half of Iran have been analyzed. Generally, the precipitation patterns reveal maximum amount of precipitation on the coasts of the Caspian Sea regions in agreement with actual precipitation data. For the Tehran region the isoprecipitation pattern shows the oscillating precipitation similar to the actual precipitation data. For obtaining more accurate results and better simulation, 4km horizontal resolution has been tried and quantitatively the precipitation values have been compared in some stations in the Tehran area. This comparison revealed that the precipitation values in most cases are more than the actual ones. In addition the comparison of thermodynamical graphs in Tehran, Mehrabad airport indicates the relative success of the model in simulating vertical profile of the atmosphere.