Avaliação da representação de bandas de nuvens tipo Zona de Convergência do Atlântico Sul (ZCAS) em modelos de previsão subsazonal
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
As previsões subsazonais consistem em prognósticos de 2 semanas a 2 meses de antecedência. Essas previsões são de grande relevância para diversos setores da sociedade, porém, os Modelos de Circulação Geral (MCGs) ainda enfrentam limitações em prever as condições atmosféricas com precisão nessa escala temporal. Este estudo teve como objetivo avaliar o desempenho de modelos de previsão subsazonal, incluindo o Modelo Brasileiro Atmosférico (BAM-1.2) do Centro de Previsão de Tempo e Estudos Climáticos do Instituto Nacional de Pesquisas Espaciais (CPTEC/INPE), na representação de bandas de nuvens, especificamente da Zona de Convergência do Atlântico Sul (ZCAS). Inicialmente, foi empregada uma metodologia automática baseada em objetos, conhecida como MetBot, para identificar e caracterizar as bandas de nuvens, através da contagem do número de dias com essas formações e identificando eventos persistentes de 4 dias ou mais, caracterizados como eventos de ZCAS. Adicionalmente, foi analisada a duração média desses eventos persistentes nas previsões numéricas em comparação com a duração observada. O desempenho das previsões foi comparado entre o modelo do CPTEC/INPE e outros quatro modelos do Projeto Sub-sazonal a Sazonal (S2S), a saber: Japan Meteorological Agency (JMA), Environment and Climate Change Canada (ECCC), European Centre for Medium Range Weather Forecasts (ECMWF) e o Australian Bureau of Meteorology (BoM). A comparação foi realizada utilizando previsões retrospectivas de 12 anos, correspondendo a previsões geradas com datas de inicialização durante os meses de novembro a março de 1999/2000 e 2010/2011, empregando métricas estatísticas como correlação linear, raiz quadrada do erro médio e estimativa dos vieses do modelo. Com isso, foi possível avaliar de forma abrangente a eficácia do modelo BAM-1.2 do CPTEC/INPE em comparação com os outros quatro modelos na representação de bandas de nuvens em escala subsazonal. Adicionalmente, este trabalho propõe a implementação de um produto para a previsão subsazonal de eventos de ZCAS, baseado no uso automatizado do algoritmo MetBot. Além disso, foi realizado um estudo de caso de uma banda de nuvens caracterizada como ZCAS, para auxiliar na avaliação comparativa do desempenho das previsões dos modelos do CPTEC/INPE e ECMWF na representação das características de precipitação, radiação de onda longa emergente no topo da atmosfera e circulação atmosférica observadas durante o caso estudado. Os resultados mostraram que, embora todos os modelos apresentem limitações, o BAM-1.2 demonstrou desempenho comparável aos modelos internacionais, com destaque para a representação de eventos persistentes. O ECMWF apresentou desempenho ligeiramente superior. Além disso, o BAM-1.2 se mostrou eficiente em prever os padrões de circulação e precipitação em eventos de ZCAS. ABSTRACT: Subseasonal forecasts involve prognostics from 2 weeks to 2 months in advance. These forecasts are of great relevance for various sectors of society, however, the General Circulation Models (GCMs) still face limitations in predicting atmospheric conditions accurately at this time scale. This study aimed to evaluate the performance of subseasonal forecast models, including the Brazilian Atmospheric Model (BAM-1.2) from the Center for Weather Forecasting and Climate Studies of the National Institute for Space Research (CPTEC/INPE), in representing cloud bands, specifically from the South Atlantic Convergence Zone (SACZ). Initially, an automatic methodology based on objects, known as MetBot, was employed to identify and characterize the cloud bands, through the counting of the number of days with these formations and identifying persistent events of 4 days or more, characterized as SACZ events. Additionally, the average duration of these persistent events in the numerical forecasts was analyzed in comparison with the observed duration. The performance of the forecasts was compared between the CPTEC/INPE model and four other models from the Subseasonal to Seasonal (S2S) Project, namely: Japan Meteorological Agency (JMA), Environment and Climate Change Canada (ECCC), European Centre for Medium Range Weather Forecasts (ECMWF) and the Australian Bureau of Meteorology (BoM). The comparison was made using retrospective forecasts of 12 years, corresponding to forecasts generated with initialization dates during the months of November to March from 1999/2000 and 2010/2011, employing statistical metrics such as linear correlation, root mean square error, and estimate of the model biases. With this, it was possible to comprehensively evaluate the efficacy of the BAM-1.2 model from CPTEC/INPE compared to the other four models in representing cloud bands on a subseasonal scale. Additionally, this work proposes the implementation of a product for the subseasonal forecast of SACZ events, based on the automated use of the MetBot algorithm. Furthermore, a case study of a cloud band characterized as SACZ was conducted to assist in the comparative evaluation of the performance of the CPTEC/INPE and ECMWF model forecasts in representing the precipitation characteristics, outgoing longwave radiation at the top of the atmosphere, and observed atmospheric circulation during the studied case. The results showed that, although all models have limitations, BAM-1.2 demonstrated performance comparable to international models, with emphasis on the representation of persistent events. The ECMWF showed slightly superior performance. In addition, BAM-1.2 proved efficient in predicting the circulation and precipitation patterns in SACZ events.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.004 | 0.003 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.005 |
| Science and technology studies | 0.002 | 0.002 |
| Scholarly communication | 0.003 | 0.002 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.003 | 0.002 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it