Impacto das parametrizações de microfísica e de camada limite planetária na simulação numérica de um evento de precipitação extrema no oeste de Santa Catarina
Data
Autores
Título da Revista
ISSN da Revista
Título de Volume
Resumo
Extremos de tempo e de clima, conforme a World Meterological Organizational (WMO), constituem eventos raros em determinado lugar e tempo, com características não usuais em termos de magnitude, localização, duração temporal ou extensão geográfica. As tempestades são eventos convectivos que atuam isoladamente ou em forma de sistemas e que podem originar uma série de fenômenos meteorológicos severos, como, precipitação intensa, granizo, tornados, ventos fortes, rajadas e relâmpagos. Os fenômenos associados à convecção intensa estão na origem da maior parte dos danos e perdas relativos a desastres devidos ao tempo atmosférico. Logo, a previsibilidade de sistemas convectivos é importante para os processos de alerta antecipados. Os modelos de previsão numérica, ao lado de sistemas de monitoramento, são as ferramentas mais adequadas a esse objetivo. Esta pesquisa fez uso de simulações numéricas usando o modelo WRF em alta resolução espacial horizontal com 2 km. O objetivo foi verificar a sensibilidade do WRF quanto às configurações dos esquemas de camada limite planetária (PBL) e de microfísica de nuvens (MP) na acuracidade da previsão de precipitação em um evento de sistema convectivo severo que ocorreu no centro e no oeste do Estado de Santa Catarina na data de 14 de agosto de 2020. Os resultados das execuções das configurações pelo WRF foram verificadas através da precipitação coletada por estações meteorológicas de superfície na região de interesse. Foram avaliados cinco esquemas de parametrização de PBL e 28 esquemas de MP e os resultados consideraram a estatística da Raiz do Erro Quadrático Médio (Root Mean Squared Error - RMSE) entre dados coletados e valores previstos. As conclusões consideraram valores de RMSE e testes de significância estatística. Os resultados indicaram que não foram encontradas diferenças estatísticas significativas entre os esquemas de MP. Contudo, esquemas de PBL apresentaram diferenças significativas entre si. Logo, diferentes esquemas de PBL têm influência sobre os resultados do modelo na acurácia da previsão de precipitação. Considerando os resultados de PBL de forma isolada, os esquemas de PBL "Yonsei University" e "Shin-Hong" foram os que obtiveram os menores valores de erros RMSE, com média para todas as configurações, de 3,28 e 3,29, respectivamente. O esquema de PBL "MRF" teve a pior performance, pois obteve média do valor de RMSE para todos os casos de 3,61. As médias de RMSE para os esquemas PBL "Mellor-Yamada Nakanishi and Niino Level 2.5" (MYNN2) e "Boulac", resultaram em 3,39 e 3,36, respectivamente. Avaliando os resultados para ambas as parametrizações de PBL e MP, as configurações de PBL-MP que obtiveram os menores erros RMSE foram "Yonsei University" (YSU)-"Jensen ISHMAEL" e "Shin-Hong"-"Purdue Lin".
Weather and climate extremes, according to the World Meterological Organization (WMO), are rare events in a given place and time, with unusual characteristics in terms of magnitude, location, duration, or geographic extent.Storms originate from deep convection and can occur isolated or grouped into systems. They cause a range of severe weather phenomena, such as heavy precipitation, hail, tornadoes, strong winds, gusts, and lightning. Phenomena associated with intense convection are responsible for most of the damage and losses related to weather-related disasters. Therefore, the predictability of convective systems is important for early warning processes. Numerical prediction models, along with monitoring systems, are the most appropriate tools for this purpose. This research used numerical simulations using the WRF model at a high horizontal spatial resolution of 2 km. The objective was to verify the sensitivity of the WRF to planetary boundary layer (PBL) and cloud microphysics (MP) scheme configurations on the accuracy of precipitation forecasting during a severe convective system that occurred in central and western Santa Catarina State on August 14, 2020. The results of the WRF configuration executions were verified using precipitation collected by surface meteorological stations in the region of interest. Five PBL parameterization schemes and 28 MP schemes were evaluated, and the results considered the statistical Root Mean Squared Error (RMSE) between collected data and predicted values. The conclusions considered RMSE values and statistical significance tests. The results indicated that no statistically significant differences were found between the MP schemes. However, PBL schemes showed significant differences. Therefore, different PBL schemes influence the model results in precipitation forecast accuracy. Considering the PBL results in isolation, the PBL schemes "Yonsei University" and "Shin-Hong" obtained the lowest RMSE error values, with an average for all configurations of 3.28 and 3.29, respectively. The PBL scheme "MRF" had the worst performance, as it obtained an average RMSE value for all cases of 3.61. The average RMSE for the PBL schemes "Mellor-Yamada Nakanishi and Niino Level 2.5" (MYNN2) and "Boulac" resulted in 3.39 and 3.36, respectively. Evaluating the results for both PBL and MP parameterizations, the PBL-MP configurations that obtained the lowest RMSE errors were "Yonsei University" (YSU)-"Jensen ISHMAEL" and "Shin-Hong"-"Purdue Lin".
Weather and climate extremes, according to the World Meterological Organization (WMO), are rare events in a given place and time, with unusual characteristics in terms of magnitude, location, duration, or geographic extent.Storms originate from deep convection and can occur isolated or grouped into systems. They cause a range of severe weather phenomena, such as heavy precipitation, hail, tornadoes, strong winds, gusts, and lightning. Phenomena associated with intense convection are responsible for most of the damage and losses related to weather-related disasters. Therefore, the predictability of convective systems is important for early warning processes. Numerical prediction models, along with monitoring systems, are the most appropriate tools for this purpose. This research used numerical simulations using the WRF model at a high horizontal spatial resolution of 2 km. The objective was to verify the sensitivity of the WRF to planetary boundary layer (PBL) and cloud microphysics (MP) scheme configurations on the accuracy of precipitation forecasting during a severe convective system that occurred in central and western Santa Catarina State on August 14, 2020. The results of the WRF configuration executions were verified using precipitation collected by surface meteorological stations in the region of interest. Five PBL parameterization schemes and 28 MP schemes were evaluated, and the results considered the statistical Root Mean Squared Error (RMSE) between collected data and predicted values. The conclusions considered RMSE values and statistical significance tests. The results indicated that no statistically significant differences were found between the MP schemes. However, PBL schemes showed significant differences. Therefore, different PBL schemes influence the model results in precipitation forecast accuracy. Considering the PBL results in isolation, the PBL schemes "Yonsei University" and "Shin-Hong" obtained the lowest RMSE error values, with an average for all configurations of 3.28 and 3.29, respectively. The PBL scheme "MRF" had the worst performance, as it obtained an average RMSE value for all cases of 3.61. The average RMSE for the PBL schemes "Mellor-Yamada Nakanishi and Niino Level 2.5" (MYNN2) and "Boulac" resulted in 3.39 and 3.36, respectively. Evaluating the results for both PBL and MP parameterizations, the PBL-MP configurations that obtained the lowest RMSE errors were "Yonsei University" (YSU)-"Jensen ISHMAEL" and "Shin-Hong"-"Purdue Lin".
Descrição
Palavras-chave
Citação
CAMILLO, Gerson Luiz. Impacto das parametrizações de microfísica e de camada limite planetária na simulação numérica de um evento de precipitação extrema no oeste de Santa Catarina. 2025. Dissertação (Mestrado Profissional em Clima e Ambiente) – Instituto Federal de Santa Catarina, Florianópolis, 2025.
