Numerical modelling of heavy rainfall event over Madeira Island in Portugal: sensitivity to different micro physical processes

• Autor(a) principal: Dasari, Hari
• Data de Publicação: 2015
• Outros Autores: Salgado, Rui
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10174/17388; https://doi.org/10.1002/joc.4243
• topic: heavy rainfall; mesoscale model; hydrometeors; Madeira

Madeira Island, Portugal, experienced intense heavy rainfall on 20 February 2010 which caused loss of more than 40 lives and great damage. This is recorded as one of the major flash flood events during the past three decades. This heavy rainfall episode is simulated using the Advanced Research Weather Research Forecasting (ARW) model at a cloud resolving grid resolution of 1 km. A nested two-way interactive four-domain configuration centred on the island and forced laterally by the 6 hourly NCEP FNL 1 ◦ data are used. Four experiments with different cloud microphysical parameterizations are performed to examine the model sensitivity to cloud microphysics on heavy rainfall simulations. Analysis of the synoptic meteorological conditions indicated that the extreme rainfall resulted due to the effect of regional orography on the prevailing large scale flow with conditionally unstable moist air. Results indicate that the model is able to reproduce the heavy rainfall with good agreement. The experiment with the Ferrier scheme produced more rainfall than the other three experiments, as well as overestimating the quantity of rainfall. Analysis of simulated cloud microphysical properties indicated that the model-produced heavy rainfall is sensitive to different cloud microphysics schemes. Model quantitative rainfall comparisons at different stations indicate that the Lin and Thompson schemes produced a more realistic simulation of heavy rainfall with relatively better correlation and higher threat score than the other two schemes. The associated dynamic and thermodynamic parameters are well produced in these two simulations.