Evaluation of weather radar rainfall estimation for nowcasting

Bibliographic Details
Main Author: Bacelar, Luiz Carlos Salgueiro Donato
Publication Date: 2018
Other Authors: Angelis, Carlos Frederico de
Format: Article
Language: por
Source: Revista Ciência e Natura (Online)
Download full: https://periodicos.ufsm.br/cienciaenatura/article/view/29995
Summary: This article presents the first evaluation of the method used by the Brazilian Centre for Monitoring and Early Warning of Natural Disasters for the short term forecast (nowcasting) of precipitation by weather radar. In this evaluation, four cases of intense rainfall in the radius of Pico do Couto radar were studied, giving the flood risk alerts sent for the municipality of Nova Friburgo, in the state of Rio de Janeiro. In relation to rain gauge data, a median under rate of 43% for the radar precipitation was observed when using the SRI (Surface Rainfall Intensity) method. After the rainfall correction, the cross-correlation extrapolation method was evaluated for the 30, 60, 90 and 120 minute forecast horizons, with accumulated precipitation every 30 minutes. The probabilities of detection (POD) obtained higher values for lower rainfall thresholds (at least 1 or 5 mm of accumulated rainfall) when compared to larger amount accumulated (20 and 30 mm). The spatial and temporal series of rainfall presented higher errors (MAE and RMSE) for 90 and 120 minutes of forecast. For all the events, an overestimate (PBIAS positive) of the predictions was observed in relation to the observed radar rain field itself.
id UFSM-23_88f81ec20f19138b99ee5b69b8a28fe3
oai_identifier_str oai:ojs.pkp.sfu.ca:article/29995
network_acronym_str UFSM-23
network_name_str Revista Ciência e Natura (Online)
repository_id_str
spelling Evaluation of weather radar rainfall estimation for nowcastingAvaliação da chuva prevista em curto prazo por radar meteorológicoNowcastingHydrometeorolog. Rainfall estimationPrecipitationRemote sensingNowcastingHidrometeorologiaEstimativa de chuvaPrecipitaçãoSensoriamento remotoThis article presents the first evaluation of the method used by the Brazilian Centre for Monitoring and Early Warning of Natural Disasters for the short term forecast (nowcasting) of precipitation by weather radar. In this evaluation, four cases of intense rainfall in the radius of Pico do Couto radar were studied, giving the flood risk alerts sent for the municipality of Nova Friburgo, in the state of Rio de Janeiro. In relation to rain gauge data, a median under rate of 43% for the radar precipitation was observed when using the SRI (Surface Rainfall Intensity) method. After the rainfall correction, the cross-correlation extrapolation method was evaluated for the 30, 60, 90 and 120 minute forecast horizons, with accumulated precipitation every 30 minutes. The probabilities of detection (POD) obtained higher values for lower rainfall thresholds (at least 1 or 5 mm of accumulated rainfall) when compared to larger amount accumulated (20 and 30 mm). The spatial and temporal series of rainfall presented higher errors (MAE and RMSE) for 90 and 120 minutes of forecast. For all the events, an overestimate (PBIAS positive) of the predictions was observed in relation to the observed radar rain field itself.Este artigo traz a primeira avaliação do método utilizado pelo Centro Nacional de Monitoramento e Alerta de Desastres Naturaispara os prognósticos em curto prazo da precipitação por radares meteorológicos. Nesta avaliação foram estudados quatro casosde chuva intensa no raio de cobertura do radar Pico do Couto, que originaram alertas de risco para o município de nova Friburgo,no estado do Rio de Janeiro. Em relação aos dados pluviométricos in situ, foi observada uma subestimativa mediana de 43% daprecipitação pelo radar ao utilizar o método SRI (Surface Rainfall Intensity). Após a correção dos campos de chuva, a metodologiapara extrapolação por cross-correlation (correlação cruzada) foi avaliada para os horizontes de previsão de 30, 60, 90 e 120minutos, com a precipitação acumulada a cada 30 minutos. As probabilidades de detecção (POD) obtiveram maiores valores paralimiares de chuva mais baixos (pelo menos 1 ou 5 mm de chuva acumulada) ao comparado com acumulados de chuva maiores(20 e 30 mm), como observado em estudos anteriores. Da mesma forma, as séries espaciais e temporais de chuva apresentarammaiores erros (MAE e RMSE) para os prognósticos de 90 e 120 minutos de previsão. Em todos os eventos foi observada umasuperestimativa (PBIAS positivo) média das previsões em relação ao próprio campo observado de chuva do radar.Universidade Federal de Santa Maria2018-03-27info:eu-repo/semantics/articleinfo:eu-repo/semantics/publishedVersionapplication/pdfhttps://periodicos.ufsm.br/cienciaenatura/article/view/2999510.5902/2179460X29995Ciência e Natura; CIÊNCIA E NATURA, V. 40, 2018; e42Ciência e Natura; CIÊNCIA E NATURA, V. 40, 2018; e422179-460X0100-8307reponame:Revista Ciência e Natura (Online)instname:Universidade Federal de Santa Maria (UFSM)instacron:UFSMporhttps://periodicos.ufsm.br/cienciaenatura/article/view/29995/pdfCopyright (c) 2018 Ciência e Naturainfo:eu-repo/semantics/openAccessBacelar, Luiz Carlos Salgueiro DonatoAngelis, Carlos Frederico de2020-04-22T18:02:35Zoai:ojs.pkp.sfu.ca:article/29995Revistahttps://periodicos.ufsm.br/cienciaenatura/indexPUBhttps://periodicos.ufsm.br/cienciaenatura/oaicienciaenatura@ufsm.br || centraldeperiodicos@ufsm.br2179-460X0100-8307opendoar:2020-04-22T18:02:35Revista Ciência e Natura (Online) - Universidade Federal de Santa Maria (UFSM)false
dc.title.none.fl_str_mv Evaluation of weather radar rainfall estimation for nowcasting
Avaliação da chuva prevista em curto prazo por radar meteorológico
title Evaluation of weather radar rainfall estimation for nowcasting
spellingShingle Evaluation of weather radar rainfall estimation for nowcasting
Bacelar, Luiz Carlos Salgueiro Donato
Nowcasting
Hydrometeorolog
. Rainfall estimation
Precipitation
Remote sensing
Nowcasting
Hidrometeorologia
Estimativa de chuva
Precipitação
Sensoriamento remoto
title_short Evaluation of weather radar rainfall estimation for nowcasting
title_full Evaluation of weather radar rainfall estimation for nowcasting
title_fullStr Evaluation of weather radar rainfall estimation for nowcasting
title_full_unstemmed Evaluation of weather radar rainfall estimation for nowcasting
title_sort Evaluation of weather radar rainfall estimation for nowcasting
author Bacelar, Luiz Carlos Salgueiro Donato
author_facet Bacelar, Luiz Carlos Salgueiro Donato
Angelis, Carlos Frederico de
author_role author
author2 Angelis, Carlos Frederico de
author2_role author
dc.contributor.author.fl_str_mv Bacelar, Luiz Carlos Salgueiro Donato
Angelis, Carlos Frederico de
dc.subject.por.fl_str_mv Nowcasting
Hydrometeorolog
. Rainfall estimation
Precipitation
Remote sensing
Nowcasting
Hidrometeorologia
Estimativa de chuva
Precipitação
Sensoriamento remoto
topic Nowcasting
Hydrometeorolog
. Rainfall estimation
Precipitation
Remote sensing
Nowcasting
Hidrometeorologia
Estimativa de chuva
Precipitação
Sensoriamento remoto
description This article presents the first evaluation of the method used by the Brazilian Centre for Monitoring and Early Warning of Natural Disasters for the short term forecast (nowcasting) of precipitation by weather radar. In this evaluation, four cases of intense rainfall in the radius of Pico do Couto radar were studied, giving the flood risk alerts sent for the municipality of Nova Friburgo, in the state of Rio de Janeiro. In relation to rain gauge data, a median under rate of 43% for the radar precipitation was observed when using the SRI (Surface Rainfall Intensity) method. After the rainfall correction, the cross-correlation extrapolation method was evaluated for the 30, 60, 90 and 120 minute forecast horizons, with accumulated precipitation every 30 minutes. The probabilities of detection (POD) obtained higher values for lower rainfall thresholds (at least 1 or 5 mm of accumulated rainfall) when compared to larger amount accumulated (20 and 30 mm). The spatial and temporal series of rainfall presented higher errors (MAE and RMSE) for 90 and 120 minutes of forecast. For all the events, an overestimate (PBIAS positive) of the predictions was observed in relation to the observed radar rain field itself.
publishDate 2018
dc.date.none.fl_str_mv 2018-03-27
dc.type.driver.fl_str_mv info:eu-repo/semantics/article
info:eu-repo/semantics/publishedVersion
format article
status_str publishedVersion
dc.identifier.uri.fl_str_mv https://periodicos.ufsm.br/cienciaenatura/article/view/29995
10.5902/2179460X29995
url https://periodicos.ufsm.br/cienciaenatura/article/view/29995
identifier_str_mv 10.5902/2179460X29995
dc.language.iso.fl_str_mv por
language por
dc.relation.none.fl_str_mv https://periodicos.ufsm.br/cienciaenatura/article/view/29995/pdf
dc.rights.driver.fl_str_mv Copyright (c) 2018 Ciência e Natura
info:eu-repo/semantics/openAccess
rights_invalid_str_mv Copyright (c) 2018 Ciência e Natura
eu_rights_str_mv openAccess
dc.format.none.fl_str_mv application/pdf
dc.publisher.none.fl_str_mv Universidade Federal de Santa Maria
publisher.none.fl_str_mv Universidade Federal de Santa Maria
dc.source.none.fl_str_mv Ciência e Natura; CIÊNCIA E NATURA, V. 40, 2018; e42
Ciência e Natura; CIÊNCIA E NATURA, V. 40, 2018; e42
2179-460X
0100-8307
reponame:Revista Ciência e Natura (Online)
instname:Universidade Federal de Santa Maria (UFSM)
instacron:UFSM
instname_str Universidade Federal de Santa Maria (UFSM)
instacron_str UFSM
institution UFSM
reponame_str Revista Ciência e Natura (Online)
collection Revista Ciência e Natura (Online)
repository.name.fl_str_mv Revista Ciência e Natura (Online) - Universidade Federal de Santa Maria (UFSM)
repository.mail.fl_str_mv cienciaenatura@ufsm.br || centraldeperiodicos@ufsm.br
_version_ 1839277883526217728