Artificial Neural Network for Data Assimilation by WRF Model in Rio de Janeiro, Brazil

2020 ◽  
Vol 38 (2) ◽  
Author(s):  
Vinícius Albuquerque de Almeida Albuquerque de Almeida ◽  
Gutemberg Borges França ◽  
Haroldo Fraga Campos Velho ◽  
Nelson F. Favilla Ebecken
Keyword(s):  
Neural Network ◽  
Relative Humidity ◽  
Data Assimilation ◽  
Air Temperature ◽  
Rio De Janeiro ◽  
Mean Squared Error ◽  
Grid Point ◽  
Wrf Model ◽  
Particle Collision ◽  
The Neural Network

ABSTRACTThis study investigates the use of neural networks for data assimilation of local data in the WRF model in Rio de Janeiro, Brazil. Surface and upper-air data (air temperature, relative humidity and wind speed and direction) from airport stations and 6-hour forecast from WRF are used as input for the model and the 3D-Var analysis for each grid point is used as target variable. Periods of 168h from 2014 and 2015 are used with 6h and 12h assimilation cycles for surface and upper-air data, respectively. The neural network model was built using the Multi-Particle Collision Algorithm (MPCA) where different topologies are tested until the optimum solution is found. Results show that the neural network is able to emulate the 3D-Var with root mean squared error (standard deviation), respectively, of 0.31 K (0.37 K), 3.10% (4.04%), 0.63 ms-1 (1.05 ms-1), 1.10 ms-1 (1.56 ms-1) for air temperature, relative humidity, u-component of the wind and v-component of the wind. Also, the results show the neural network method is able to run 71 times faster than the conventional method under similar hardware configurations.RESUMOEste estudo investiga o uso de redes neurais para assimilação de dados locais no modelo WRF no Rio de Janeiro. Dados de superfície e do ar superior (temperatura do ar, umidade relativa e velocidade e direção do vento) das estações do aeroporto e previsão de 6 horas do WRF são usados como entrada para o modelo, e a análise 3D-Var para cada ponto da grade é usada como variável destino. Períodos de 168h de 2014 e 2015 são utilizados com ciclos de assimilação de 6h e 12h para dados de superfície e do ar superior, respectivamente. O modelo de rede neural foi construído usando o algoritmo de colisão de partículas múltiplas (MPCA), onde diferentes topologias são testadas até que a solução ideal seja encontrada. Os resultados mostram que a rede neural é capaz de emular o 3D-Var com raiz do erro quadrático médio (desvio padrão) de 0,31 K (0,37 K), 3,10% (4,04%), 0,63 ms -1 (1,05 ms-1), 1,10 ms-1 (1,56 ms-1) para temperatura do ar, umidade relativa, componente u do vento e componente v do vento. Além disso, os resultados mostram que o método de rede neural é capaz de rodar 71 vezes mais rápido que o método convencional em configurações de hardware semelhantes.

scholarly journals Artificial Neural Network for Data Assimilation by WRF Model in Rio de Janeiro, Brazil

2020 ◽  
Vol 38 (2) ◽  
Author(s):  
Vinícius Albuquerque de Almeida Albuquerque de Almeida ◽  
Gutemberg Borges França ◽  
Haroldo Fraga Campos Velho ◽  
Nelson F. Favilla Ebecken
Keyword(s):  
Neural Network ◽  
Relative Humidity ◽  
Data Assimilation ◽  
Air Temperature ◽  
Rio De Janeiro ◽  
Mean Squared Error ◽  
Grid Point ◽  
Wrf Model ◽  
Particle Collision ◽  
The Neural Network

ABSTRACTThis study investigates the use of neural networks for data assimilation of local data in the WRF model in Rio de Janeiro, Brazil. Surface and upper-air data (air temperature, relative humidity and wind speed and direction) from airport stations and 6-hour forecast from WRF are used as input for the model and the 3D-Var analysis for each grid point is used as target variable. Periods of 168h from 2014 and 2015 are used with 6h and 12h assimilation cycles for surface and upper-air data, respectively. The neural network model was built using the Multi-Particle Collision Algorithm (MPCA) where different topologies are tested until the optimum solution is found. Results show that the neural network is able to emulate the 3D-Var with root mean squared error (standard deviation), respectively, of 0.31 K (0.37 K), 3.10% (4.04%), 0.63 ms-1 (1.05 ms-1), 1.10 ms-1 (1.56 ms-1) for air temperature, relative humidity, u-component of the wind and v-component of the wind. Also, the results show the neural network method is able to run 71 times faster than the conventional method under similar hardware configurations. Redes Neurais Artificiais para Assimilação de Dados no Modelo WRF no Rio de Janeiro, BrazilRESUMOEste estudo investiga o uso de redes neurais para assimilação de dados locais no modelo WRF no Rio de Janeiro. Dados de superfície e do ar superior (temperatura do ar, umidade relativa e velocidade e direção do vento) das estações do aeroporto e previsão de 6 horas do WRF são usados como entrada para o modelo, e a análise 3D-Var para cada ponto da grade é usada como variável destino. Períodos de 168h de 2014 e 2015 são utilizados com ciclos de assimilação de 6h e 12h para dados de superfície e do ar superior, respectivamente. O modelo de rede neural foi construído usando o algoritmo de colisão de partículas múltiplas (MPCA), onde diferentes topologias são testadas até que a solução ideal seja encontrada. Os resultados mostram que a rede neural é capaz de emular o 3D-Var com raiz do erro quadrático médio (desvio padrão) de 0,31 K (0,37 K), 3,10% (4,04%), 0,63 ms -1 (1,05 ms -1), 1,10 ms -1 (1,56 ms -1) para temperatura do ar, umidade relativa, componente u do vento e componente v do vento. Além disso, os resultados mostram que o método de rede neural é capaz de rodar 71 vezes mais rápido que o método convencional em configurações de hardware semelhantes.

Artificial Neural Network for Data Assimilation by WRF Model in Rio de Janeiro, Brazil

2020 ◽  
Vol 38 (2) ◽  
Author(s):  
Vinícius Albuquerque de Almeida Albuquerque de Almeida ◽  
Gutemberg Borges França ◽  
Haroldo Fraga Campos Velho ◽  
Nelson F. Favilla Ebecken
Keyword(s):  
Neural Network ◽  
Relative Humidity ◽  
Data Assimilation ◽  
Air Temperature ◽  
Rio De Janeiro ◽  
Mean Squared Error ◽  
Grid Point ◽  
Wrf Model ◽  
Particle Collision ◽  
The Neural Network

ABSTRACTThis study investigates the use of neural networks for data assimilation of local data in the WRF model in Rio de Janeiro, Brazil. Surface and upper-air data (air temperature, relative humidity and wind speed and direction) from airport stations and 6-hour forecast from WRF are used as input for the model and the 3D-Var analysis for each grid point is used as target variable. Periods of 168h from 2014 and 2015 are used with 6h and 12h assimilation cycles for surface and upper-air data, respectively. The neural network model was built using the Multi-Particle Collision Algorithm (MPCA) where different topologies are tested until the optimum solution is found. Results show that the neural network is able to emulate the 3D-Var with root mean squared error (standard deviation), respectively, of 0.31 K (0.37 K), 3.10% (4.04%), 0.63 ms-1 (1.05 ms-1), 1.10 ms-1 (1.56 ms-1) for air temperature, relative humidity, u-component of the wind and v-component of the wind. Also, the results show the neural network method is able to run 71 times faster than the conventional method under similar hardware configurations. Redes Neurais Artificiais para Assimilação de Dados no Modelo WRF no Rio de Janeiro, BrazilRESUMOEste estudo investiga o uso de redes neurais para assimilação de dados locais no modelo WRF no Rio de Janeiro. Dados de superfície e do ar superior (temperatura do ar, umidade relativa e velocidade e direção do vento) das estações do aeroporto e previsão de 6 horas do WRF são usados como entrada para o modelo, e a análise 3D-Var para cada ponto da grade é usada como variável destino. Períodos de 168h de 2014 e 2015 são utilizados com ciclos de assimilação de 6h e 12h para dados de superfície e do ar superior, respectivamente. O modelo de rede neural foi construído usando o algoritmo de colisão de partículas múltiplas (MPCA), onde diferentes topologias são testadas até que a solução ideal seja encontrada. Os resultados mostram que a rede neural é capaz de emular o 3D-Var com raiz do erro quadrático médio (desvio padrão) de 0,31 K (0,37 K), 3,10% (4,04%), 0,63 ms -1 (1,05 ms -1), 1,10 ms -1 (1,56 ms -1) para temperatura do ar, umidade relativa, componente u do vento e componente v do vento. Além disso, os resultados mostram que o método de rede neural é capaz de rodar 71 vezes mais rápido que o método convencional em configurações de hardware semelhantes.

Artificial Neural Network for Data Assimilation by WRF Model in Rio de Janeiro, Brazil

2020 ◽  
Vol 38 (2) ◽  
Author(s):  
Vinícius Albuquerque de Almeida Albuquerque de Almeida ◽  
Gutemberg Borges França ◽  
Haroldo Fraga Campos Velho ◽  
Nelson F. Favilla Ebecken
Keyword(s):  
Neural Network ◽  
Relative Humidity ◽  
Data Assimilation ◽  
Air Temperature ◽  
Rio De Janeiro ◽  
Mean Squared Error ◽  
Grid Point ◽  
Wrf Model ◽  
Particle Collision ◽  
The Neural Network

ABSTRACTThis study investigates the use of neural networks for data assimilation of local data in the WRF model in Rio de Janeiro, Brazil. Surface and upper-air data (air temperature, relative humidity and wind speed and direction) from airport stations and 6-hour forecast from WRF are used as input for the model and the 3D-Var analysis for each grid point is used as target variable. Periods of 168h from 2014 and 2015 are used with 6h and 12h assimilation cycles for surface and upper-air data, respectively. The neural network model was built using the Multi-Particle Collision Algorithm (MPCA) where different topologies are tested until the optimum solution is found. Results show that the neural network is able to emulate the 3D-Var with root mean squared error (standard deviation), respectively, of 0.31 K (0.37 K), 3.10% (4.04%), 0.63 ms-1 (1.05 ms-1), 1.10 ms-1 (1.56 ms-1) for air temperature, relative humidity, u-component of the wind and v-component of the wind. Also, the results show the neural network method is able to run 71 times faster than the conventional method under similar hardware configurations. Redes Neurais Artificiais para Assimilação de Dados no Modelo WRF no Rio de Janeiro, BrazilRESUMOEste estudo investiga o uso de redes neurais para assimilação de dados locais no modelo WRF no Rio de Janeiro. Dados de superfície e do ar superior (temperatura do ar, umidade relativa e velocidade e direção do vento) das estações do aeroporto e previsão de 6 horas do WRF são usados como entrada para o modelo, e a análise 3D-Var para cada ponto da grade é usada como variável destino. Períodos de 168h de 2014 e 2015 são utilizados com ciclos de assimilação de 6h e 12h para dados de superfície e do ar superior, respectivamente. O modelo de rede neural foi construído usando o algoritmo de colisão de partículas múltiplas (MPCA), onde diferentes topologias são testadas até que a solução ideal seja encontrada. Os resultados mostram que a rede neural é capaz de emular o 3D-Var com raiz do erro quadrático médio (desvio padrão) de 0,31 K (0,37 K), 3,10% (4,04%), 0,63 ms -1 (1,05 ms -1), 1,10 ms -1 (1,56 ms -1) para temperatura do ar, umidade relativa, componente u do vento e componente v do vento. Além disso, os resultados mostram que o método de rede neural é capaz de rodar 71 vezes mais rápido que o método convencional em configurações de hardware semelhantes.

Data Assimilation Using WRFDA Over the Terminal Area of Rio de Janeiro

2021 ◽  
Vol 36 (1) ◽  
pp. 87-96
Author(s):  
Vinícius Albuquerque de Almeida ◽  
Gutemberg Borges França ◽  
Haroldo Fraga de Campos Velho ◽  
Nelson Francisco Favilla Ebecken
Keyword(s):  
Relative Humidity ◽  
Data Assimilation ◽  
Rio De Janeiro ◽  
Wrf Model ◽  
Observation Data ◽  
Local Conditions ◽  
Forecast Data ◽  
Assimilation Process ◽  
The Impact ◽  
Terminal Area

Abstract The impact of the data assimilation process of air temperature and relative humidity from surface meteorological stations and sounding at airports in the terminal area of Rio de Janeiro is evaluated using the Weather Research and Forecast Data Assimilation system. Synthetic data of temperature, relative humidity and wind are generated in the locations of airport sensors by applying a white-noise perturbation in the forecast data. Results show a positive overall impact of the assimilation process with the removal of part of the noise in the observation data but keeping the effect of local conditions in the later timesteps of the simulation. In addition, with the assimilation process there is a global reduction of the error between the analysis data and the observation data. In the future, a neural network will be trained to emulate the data assimilation process to speed-up the assimilation process in the WRF model.

scholarly journals Effect of meteorological factors on photovoltaic power forecast based on the neural network

RSC Advances ◽  
2017 ◽  
Vol 7 (88) ◽  
pp. 55846-55850 ◽  
Author(s):  
Wenbo Xiao ◽  
Jin Dai ◽  
Huaming Wu ◽  
Gina Nazario ◽  
Feng Cheng
Keyword(s):  
Neural Network ◽  
Wind Speed ◽  
Relative Humidity ◽  
Air Temperature ◽  
Meteorological Factors ◽  
Network Models ◽  
Neural Network Models ◽  
Photovoltaic Power ◽  
The Neural Network

In this paper, the effects of meteorological factors (including air temperature, wind speed, and relative humidity) on photovoltaic (PV) power forecast using neural network models have been studied.

Tunicate swarm algorithm-trained multi-layered perceptron for data centre energy demand forecasting and relative percentage contribution analysis of input parameters

2021 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Oluwafemi Ajayi ◽  
Reolyn Heymann
Keyword(s):  
Neural Network ◽  
Energy Management ◽  
Energy Demand ◽  
Mean Squared Error ◽  
Data Set ◽  
Content Type ◽  
Demand Pattern ◽  
The Neural Network ◽  
Input Parameters ◽  
Demand Profile

Purpose Energy management is critical to data centres (DCs) majorly because they are high energy-consuming facilities and demand for their services continue to rise due to rapidly increasing global demand for cloud services and other technological services. This projected sectoral growth is expected to translate into increased energy demand from the sector, which is already considered a major energy consumer unless innovative steps are used to drive effective energy management systems. The purpose of this study is to provide insights into the expected energy demand of the DC and the impact each measured parameter has on the building's energy demand profile. This serves as a basis for the design of an effective energy management system. Design/methodology/approach This study proposes novel tunicate swarm algorithm (TSA) for training an artificial neural network model used for predicting the energy demand of a DC. The objective is to find the optimal weights and biases of the model while avoiding commonly faced challenges when using the backpropagation algorithm. The model implementation is based on historical energy consumption data of an anonymous DC operator in Cape Town, South Africa. The data set provided consists of variables such as ambient temperature, ambient relative humidity, chiller output temperature and computer room air conditioning air supply temperature, which serve as inputs to the neural network that is designed to predict the DC’s hourly energy consumption for July 2020. Upon preprocessing of the data set, total sample number for each represented variable was 464. The 80:20 splitting ratio was used to divide the data set into training and testing set respectively, making 452 samples for the training set and 112 samples for the testing set. A weights-based approach has also been used to analyze the relative impact of the model’s input parameters on the DC’s energy demand pattern. Findings The performance of the proposed model has been compared with those of neural network models trained using state of the art algorithms such as moth flame optimization, whale optimization algorithm and ant lion optimizer. From analysis, it was found that the proposed TSA outperformed the other methods in training the model based on their mean squared error, root mean squared error, mean absolute error, mean absolute percentage error and prediction accuracy. Analyzing the relative percentage contribution of the model's input parameters based on the weights of the neural network also shows that the ambient temperature of the DC has the highest impact on the building’s energy demand pattern. Research limitations/implications The proposed novel model can be applied to solving other complex engineering problems such as regression and classification. The methodology for optimizing the multi-layered perceptron neural network can also be further applied to other forms of neural networks for improved performance. Practical implications Based on the forecasted energy demand of the DC and an understanding of how the input parameters impact the building's energy demand pattern, neural networks can be deployed to optimize the cooling systems of the DC for reduced energy cost. Originality/value The use of TSA for optimizing the weights and biases of a neural network is a novel study. The application context of this study which is DCs is quite untapped in the literature, leaving many gaps for further research. The proposed prediction model can be further applied to other regression tasks and classification tasks. Another contribution of this study is the analysis of the neural network's input parameters, which provides insight into the level to which each parameter influences the DC’s energy demand profile.

scholarly journals A Heuristic Approach for Precipitation Data Assimilation: Characterization Using OSSEs

2019 ◽  
Vol 147 (9) ◽  
pp. 3445-3466 ◽  
Author(s):  
Andrés A. Pérez Hortal ◽  
Isztar Zawadzki ◽  
M. K. Yau
Keyword(s):  
Data Assimilation ◽  
Initial Conditions ◽  
Grid Point ◽  
Wrf Model ◽  
Ensemble Member ◽  
Absolute Difference ◽  
Time Interval ◽  
State Variables ◽  
Vertical Column ◽  
A New Technique

Abstract We introduce a new technique for the assimilation of precipitation observations, the localized ensemble mosaic assimilation (LEMA). The method constructs an analysis by selecting, for each vertical column in the model, the ensemble member with precipitation at the ground that is locally closest to the observed values. The proximity between the modeled and observed precipitation is determined by the mean absolute difference of precipitation intensity, converted to reflectivity and measured over a spatiotemporal window centered at each grid point of the model. The underlying hypothesis of the approach is that the ensemble members that are locally closer to the observed precipitation are more probable to be closer to the “truth” in the state variables than the other members. The initial conditions for the new forecast are obtained by nudging the background states toward the mosaic of the closest ensemble members (analysis) over a 30 min time interval, reducing the impacts of the imbalances at the boundaries between the different selected members. The potential of the method is studied using observing system simulation experiments (OSSEs) employing a small ensemble of 20 members. The ensemble is produced by the WRF Model, run at a horizontal grid spacing of 20 km. The experiments lend support to the validity of the hypothesis and allow the determination of the optimal parameters for the approach. In the context of OSSE, this new data assimilation technique is able to produce forecasts with considerable and long-lived error reductions in the fields of precipitation, temperature, humidity, and wind.

scholarly journals WRF-Chem Simulation of Winter Visibility in Jiangsu, China, and the Application of a Neural Network Algorithm

Atmosphere ◽  
2020 ◽  
Vol 11 (5) ◽  
pp. 520
Author(s):  
Peishu Zong ◽  
Yali Zhu ◽  
Huijun Wang ◽  
Duanyang Liu
Keyword(s):  
Neural Network ◽  
Wind Speed ◽  
Extinction Coefficient ◽  
Wrf Model ◽  
Jiangsu Province ◽  
Network Algorithm ◽  
The Neural Network ◽  
Noticeable Improvement ◽  
Neural Network Algorithm ◽  
Meteorological Elements

In this paper, the winter visibility in Jiangsu Province is simulated by WRF-Chem (Weather Research and Forecasting (WRF) model coupled with Chemistry) with high spatiotemporal resolutions. Simulation results show that WRF-Chem has good capability to simulate the visibility and related local meteorological elements and air pollutants in Jiangsu in the winters of 2013–2017. For visibility inversion, this study adopts the neural network algorithm. Meteorological elements, including wind speed, humidity and temperature, are introduced to improve the performance of WRF-Chem relative to the visibility inversion scheme, which is based on the Interagency Monitoring of Protected Visual Environments (IMPROVE) extinction coefficient algorithm. The neural network offers a noticeable improvement relative to the inversion scheme of the IMPROVE visibility extinction coefficient, substantially improving the underestimation of winter visibility in Jiangsu Province. For instance, the correlation coefficient increased from 0.17 to 0.42, and root mean square error decreased from 2.62 to 1.76. The visibility inversion results under different humidity and wind speed levels show that the underestimation of the visibility using the IMPROVE scheme is especially remarkable. However, the underestimation issue is essentially solved using the neural network algorithm. This study serves as a basis for further predicting winter haze events in Jiangsu Province using WRF-Chem and deep-learning methods.

scholarly journals Computational Fluid Dynamic Design and Experimental Study of a Temperature Sensor Array Used in Climate Reference Station

2019 ◽  
Vol 36 (9) ◽  
pp. 1835-1847
Author(s):  
Jie Yang ◽  
Qingquan Liu ◽  
Wei Dai
Keyword(s):  
Neural Network ◽  
Network Model ◽  
Air Temperature ◽  
Neural Network Model ◽  
Temperature Sensor ◽  
Sensor Array ◽  
The Neural Network ◽  
Temperature Observation ◽  
The Mean ◽  
Cfd Method

Accurate air temperature measurements are demanded for climate change research. However, air temperature sensors installed in a screen or a radiation shield have traditionally resisted observation accuracy due to a number of factors, particularly solar radiation. Here we present a novel temperature sensor array to improve the air temperature observation accuracy. To obtain an optimum design of the sensor array, we perform a series of analyses of the sensor array with various structures based on a computational fluid dynamics (CFD) method. Then the CFD method is applied to obtain quantitative radiation errors of the optimum temperature sensor array. For further improving the measurement accuracy of the sensor array, an artificial neural network model is developed to learn the relationship between the radiation error and environment variables. To assess the extent to which the actual performance adheres to the theoretical CFD model and the neural network model, air temperature observation experiments are conducted. An aspirated temperature measurement platform with a forced airflow rate up to 20 m s−1 served as an air temperature reference. The average radiation errors of a temperature sensor equipped with a naturally ventilated radiation shield and a temperature sensor installed in a screen are 0.42° and 0.23°C, respectively. By contrast, the mean radiation error of the temperature sensor array is approximately 0.03°C. The mean absolute error (MAE) between the radiation errors provided by the experiments and the radiation errors given by the neural network model is 0.007°C, and the root-mean-square error (RMSE) is 0.009°C.

Sign in / Sign up

Export Citation Format

Share Document