Many natural disasters in South America are linked to meteorological phenomena. Therefore, forecasting and monitoring climatic events are fundamental issues for society and various sectors of the economy. In the last decades, machine learning models have been developed to tackle different issues in society, but there is still a gap in applications to applied physics. Here, different machine learning models are evaluated for precipitation prediction over South America. Currently, numerical weather prediction models are unable to precisely reproduce the precipitation patterns in South America due to many factors such as the lack of region-specific parametrizations and data availability. The results are compared to the general circulation atmospheric model currently used operationally in the National Institute for Space Research (INPE: Instituto Nacional de Pesquisas Espaciais), Brazil. Machine learning models are able to produce predictions with errors under 2 mm in most of the continent in comparison to satellite-observed precipitation patterns for different climate seasons, and also outperform INPE’s model for some regions (e.g., reduction of errors from 8 to 2 mm in central South America in winter). Another advantage is the computational performance from machine learning models, running faster with much lower computer resources than models based on differential equations currently used in operational centers. Therefore, it is important to consider machine learning models for precipitation forecasts in operational centers as a way to improve forecast quality and to reduce computation costs.
The process of data assimilation, in which meteorological observations and weather forecasts are merged to provide an analysis field, has been largely studied by the scientific community and operational centers. The 3D-Variational (3D-Var) approach available in the Weather Research and Forecast (WRF) computer model is evaluated for data assimilation for the Terminal Control Area of Rio de Janeiro (TCA-RJ). The basic goal of any variational data assimilation system is to produce an optimal estimate of the atmospheric state at analysis time. The analysis field is estimated from a first guess (previous forecast) and an observation field, weighted by the error matrices. The WRF is designed for nowcasting (forecasts up to 6h) for the TCA-RJ through assimilation cycles using surface, sounding, and wind profile data. The preliminary results show the model sensibility for each observation type and encourage the use of this technique operationally for the support of the air traffic management controlled by the Brazilian Air Force.
El sistema del monzón sudamericano (SAMS, por sus siglas en inglés) se refiere a la fuerte variabilidad estacional de la precipitación observada en América del Sur, con lluvias intensas durante el verano, mientras que el invierno es la estación más seca. El objetivo de este trabajo es explicar cómo los diferentes conjuntos de datos de precipitación de reanálisis global (NCEP-2, ERA-Interim y CFSR-1) representan los patrones de lluvia y su intensidad asociados con el SAMS en comparación con el análisis basado en medidores (CPC). Los resultados muestran que los productos de reanálisis tienen algunas dificultades para simular tanto la distribución espacial como la intensidad de la precipitación observada. Sin embargo, ERA-Interim parece correlacionarse mejor con la variabilidad observada en la precipitación durante la temporada cálida austral. La intensidad y el posicionamiento de los flujos de humedad a 850 hPa y los patrones de divergencia asociados podrían explicar las diferencias encontradas en la precipitación de tres productos de reanálisis global en la región central del SAMS.
ABSTRACTThe South American monsoon system (SAMS) refers to the strong seasonal variability in precipitation observed over South America, with high precipitation amounts during the summer while winter is the driest season. The aim of this work is to understand how different global reanalysis precipitation datasets (NCEP-2, ERA-Interim and CFSR-1) represent rainfall patterns and intensity associated with SAMS in comparison to gauge-based analysis (CPC). The results show that reanalysis products have some difficulties in simulating both spatial distribution and intensity of the observed precipitation. However, ERA-Interim seems to better correlate with the observed variability in precipitation during the austral warm season. The intensity and positioning of the 850-hPa moisture fluxes and the associated divergence patterns might explain differences found in precipitation from the three global reanalysis products in the SAMS core region.
Abstract. The practical feasibility of neural networks models for data assimilation using local observations data in the WRF model for the Rio de Janeiro metropolitan region in Brazil is evaluated. Surface and multi-level variables retrieved from airport meteorological stations are used: air temperature, relative humidity, and wind (speed and direction). Also, 6-hour forecast from WRF high-resolution simulations are used – domain centered in the Rio de Janeiro city with nested grids of 8 and 2.6 km. Periods of 168 h from 2015–2019 are used with 6 h and 12 h assimilation cycles for surface and upper-air data, respectively, applied to 6-hour forecast fields. The observed data (interpolated to grid points close to airport locations and influence computed in its surroundings) and short-range forecasts are used as input for training model and the 3D-Var analysis on 6-hour forecast fields for each grid point is used as target variable. The neural network models are built using two different approaches: WEKA mul- tilayer perceptron model and TensorFlow’s deep learning implementation. The year of 2019 is used as an independent dataset for forecast validation from the trained models. Results employing 6-hour forecast fields with neural network models are able to emulate the 3D-Var results for surface and multi-level variables, with better results for the NN-TensoFlow implementation. The main result refers to CPU time reduction enabled by the neural networks models, reducing the data assimilation CPU-time by 121 times and 25 times for NN-TensorFlow and NN-WEKA, respectively, in comparison to the 3D-Var method under the same hardware configurations.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.