Exploring the ability of the WRF-ARW atmospheric model to simulate different meteorological conditions in Greece

摘要

Studies on climate change reveal as its main consequence the mean global temperature rise, as well as the increase of extreme weather events frequency and intensity. The latter are more frequently occurring worldwide over the last years and are becoming of great concern due to their strong societal and economic impacts. The best possible simulation of such events resembling reality and hence, the improvement of their predictability are very important milestones to the scientific community so as to limit down the impact of the related hazards. The WRF-ARW atmospheric model was employed to simulate different meteorological conditions in the framework of compiling a computational approach to forecast and study extremes for risk assessment at country level and particularly in Greece. Firstly, the extra-tropical cyclone (Cleopatra), in November 2013, was investigated with a number of sensitivity tests: a "chemistry including setup" of the WRF model, then the same configuration with the "chemistry" option deactivated and finally, a 3-nest setup with 1-way and 2-way nesting. The follow up of these studies provided the optimum model setup that was the 1-way nesting, which was used afterwards to simulate an extreme hot weather event that occurred in June 2007. Moreover, we examined the performance of this resolved WRF setup for a usual low-pressure system that affected Greece in September 2015. In these studies, NCEP FNL analyses data were used for model input. The main meteorological variables were analyzed and the evaluation of the results was performed with comparisons against in-situ measurements by the network of the Hellenic National Meteorological Service (HNMS) stations. The comparisons showed, in general, very good agreement between the model temperature fields and observations in all test cases. The 3-nest setup with 1-way nesting approach provided high quality prediction of the temperature fields locally at high resolution in the cases of the simulated extreme events. The model behaviour differed between the simulations of the two extreme events and the usual-low pressure system in the calculation of the precipitation variable, particularly over the areas of complex topography. Overall, the model configuration of 1-way nesting proved to be a useful modelling approach for studying and forecasting such events in an attempt to limit down their consequences.