P50 WRF simulations of extreme rainfall cases in Manaus, Brazil: Assessing the performances of cumulus and microphysics schemes

Veiga, José, Rita Souza, Priscila Miranda, Erilane Teixeira, University of Amazonas State, Luiz Cândido, National Institute for Amazonian Research, and Julio Tota, Federal University of Western Pará

The city of Manaus is frequently hit by mesoscale weather phenomenas that cause diverse kinds of problems for citizens, such as, electrical energy loss, massive flooding, landslides, and terrestrial vehicle and air traffic disruption. For this reason, we believe that better forecasts should be provided for the citizens of Manaus in order to minimize impacts caused by severe weather phenomenas. To evaluate the performance of WRF, we selected two squall line events that struck Manaus on 11th and 30th September, 2013. These events caused huge quantities of rainfall to accumulate in the city. One of the meteorological surface stations recorded 104 mm of rainfall in less than an hour (where 76 mm is the expected rainfall quantity for September). Several numerical experiments with WRF were performed to evaluate 3 convective and 11 microphysics parameterisations. The simulation time for each experiment was of 40 hours. Four domains, 27, 9, 3 and 1 km of horizontal resolution and 28 model levels were used. The comparisons between model output and observed data were based on 5 urban meteorological stations. Regardless of meteorological variables, results show that Kain-Fritsch cumulus parameterisation performed better than other cumulus schemes, while microphysical schemes are dependent on meteorological variables.