Study on Wind Simulations Using Deep Learning Techniques during Typhoons: A Case Study of Northern Taiwan

A scheme for wind-speed simulation during typhoons in Taiwan is highly desirable, considering the effects of the powerful winds accompanying the severe typhoons. The developed combination of deep learning (DL) algorithms with a weather-forecasting numerical model can be used to determine wind speed in a rapid simulation process. Here, the Weather Research and Forecasting (WRF) numerical model was employed as the numerical simulation-based model for precomputing solutions to determine the wind velocity at arbitrary positions where the wind cannot be measured. The deep neural network (DNN) was used for constructing the DL-based wind-velocity simulation model. The experimental area of Northern Taiwan was used for the simulation. Regarding the complex typhoon system, the collected data comprised the typhoon tracks, FNL (Final) Operational Global Analysis Data for the WRF model, typhoon characteristics, and ground weather data. This study included 47 typhoon events that occurred over 2000–2017. Three measures were used to analyze the models for identifying optimal performance levels: Mean absolute error, root mean squared error, and correlation coefficient. This study compared observations with the WRF numerical model and DNN model. The results revealed that (1) simulations by using the WRF-based models were satisfactorily consistent with the observed data and (2) simulations by using the DNN model were considerably consistent with those of the WRF-based model. Consequently, the proposed DNN combined with WRF model can be effectively used in simulations of wind velocity at arbitrary positions of study area.

Download Full-text

Prediction of Wind Speed, Potential Wind Power, and the Associated Uncertainties for Offshore Wind Farm Using Deep Learning

ASME 2020 Power Conference ◽

10.1115/power2020-16557 ◽

2020 ◽

Author(s):

Do-Eun Choe ◽

Gary Talor ◽

Changkyu Kim

Keyword(s):

Power Generation ◽

Deep Learning ◽

Wind Speed ◽

Wind Power ◽

Wind Power Generation ◽

Offshore Wind ◽

Weather Data ◽

Local Wind ◽

Offshore Wind Power ◽

The Future

Abstract Floating offshore wind turbines hold great potential for future solutions to the growing demand for renewable energy production. Thereafter, the prediction of the offshore wind power generation became critical in locating and designing wind farms and turbines. The purpose of this research is to improve the prediction of the offshore wind power generation by the prediction of local wind speed using a Deep Learning technique. In this paper, the future local wind speed is predicted based on the historical weather data collected from National Oceanic and Atmospheric Administration. Then, the prediction of the wind power generation is performed using the traditional methods using the future wind speed data predicted using Deep Learning. The network layers are designed using both Long Short-Term Memory (LSTM) and Bi-directional LSTM (BLSTM), known to be effective on capturing long-term time-dependency. The selected networks are fine-tuned, trained using a part of the weather data, and tested using the other part of the data. To evaluate the performance of the networks, a parameter study has been performed to find the relationships among: length of the training data, prediction accuracy, and length of the future prediction that is reliable given desired prediction accuracy and the training size.

Download Full-text

Evaluation of the WRF-UCM mesoscale model and ECMWF global operational forecasts over the Paris region in the prospect of tracer atmospheric transport modeling

Elem Sci Anth ◽

10.1525/elementa.319 ◽

2018 ◽

Vol 6 ◽

Cited By ~ 3

Author(s):

Jinghui Lian ◽

Lin Wu ◽

François-Marie Bréon ◽

Grégoire Broquet ◽

Robert Vautard ◽

...

Keyword(s):

Wind Speed ◽

Urban Areas ◽

Mesoscale Model ◽

Atmospheric Transport ◽

Research Question ◽

Wrf Model ◽

Co2 Flux ◽

Weather Data ◽

Energy Budgets ◽

Atmospheric Measurements

The quantification of CO2 emissions from cities using atmospheric measurements requires accurate knowledge of the atmospheric transport. Complex urban terrains significantly modify surface roughness, augment surface energy budgets, and create heat islands, all of which lead to lower horizontal winds and enhanced convection over urban areas. The question remains whether these processes should be included in atmospheric transport models that are used for city scale CO2 inversion, and whether they need to be tailored on a city basis. In this study, we use the WRF model over Paris to address the following research question: does WRF runs at a 3 km resolution, including urban effects and the assimilation of local weather data, perform better than ECMWF forecasts that give fields at 16 km resolution? The analysis of model performances focuses on three variables: air temperature, wind and the planetary boundary layer (PBL) height. The results show that the use of objective analysis and nudging tools are required to obtain good agreements between WRF simulated fields with observations. Surface temperature is well reproduced by both WRF and ECMWF forecasts, with correlation coefficients with hourly observations larger than 0.92 and MBEs within 1°C over one month. Wind speed correlations with hourly observations are similar for WRF (range 0.76~0.85 across stations) and ECMWF (0.79~0.84), but the associated RMSEs and MBEs are better for ECMWF. Conversely, WRF outperforms ECMWF forecasts for its description of wind direction, horizontal and vertical gradients. Sensitivity tests with different WRF physics schemes show that the wind speed and the PBL height are strongly influenced by PBL schemes. The marginal advantage of WRF over ECMWF for the desired application is sufficient to motive additional testing with prescribed CO2 flux maps for comparing modeled CO2 concentrations with available observations in an urban environment.

Download Full-text

Avaliação do desempenho do Modelo WRF para prognóstico do vento na Região Central de Alagoas – Craíbas

Ciência e Natura ◽

10.5902/2179460x30759 ◽

2018 ◽

Vol 40 ◽

pp. 187

Author(s):

Silvania Maria Santos da Silva ◽

Roberto Fernando da Fonseca Lyra ◽

Rosiberto Salustiano da Silva Júnior ◽

Nareida Simone Delgado da Cruz ◽

Sâmara Dos Santos Silva

Keyword(s):

Wind Speed ◽

Wind Velocity ◽

Central Region ◽

Wrf Model ◽

Wind Farms ◽

Dry Season ◽

Mean Difference ◽

Good Tool ◽

The Mean ◽

Brazilian Northeast

The aim of this work was to evaluate the performance of the WRF model for estimate wind speed in the central region of Alagoas State (Brazil). The wind velocity (WRF outputs) were compared with anemometric data of the PVPN projetc (Previsão do Vento em Parques Eólicos no Nordeste Brasileiro “Wind forecast for wind farms in the Brazilian Northeast”), from January to December 2014. The results showed that the WRF model estimated very well the monthly and daily wind speed averages. Deviations are greater in the dry season. Throughout the year the mean difference (WRF-OBS) was only 0.23 m.s-1 (6.14 m.s-1 versus 5.91 m.s-1), showing that this model is a good tool for forecasting wind for wind farms.

Download Full-text

Auto-Colorization of Historical Images Using Deep Convolutional Neural Networks

Mathematics ◽

10.3390/math8122258 ◽

2020 ◽

Vol 8 (12) ◽

pp. 2258

Author(s):

Madhab Raj Joshi ◽

Lewis Nkenyereye ◽

Gyanendra Prasad Joshi ◽

S. M. Riazul Islam ◽

Mohammad Abdullah-Al-Wadud ◽

...

Keyword(s):

Neural Network ◽

Deep Learning ◽

User Study ◽

Mean Squared Error ◽

Color Image ◽

Machine Learning Techniques ◽

Global Features ◽

Black And White ◽

Historical Images ◽

Learning Techniques

Enhancement of Cultural Heritage such as historical images is very crucial to safeguard the diversity of cultures. Automated colorization of black and white images has been subject to extensive research through computer vision and machine learning techniques. Our research addresses the problem of generating a plausible colored photograph of ancient, historically black, and white images of Nepal using deep learning techniques without direct human intervention. Motivated by the recent success of deep learning techniques in image processing, a feed-forward, deep Convolutional Neural Network (CNN) in combination with Inception- ResnetV2 is being trained by sets of sample images using back-propagation to recognize the pattern in RGB and grayscale values. The trained neural network is then used to predict two a* and b* chroma channels given grayscale, L channel of test images. CNN vividly colorizes images with the help of the fusion layer accounting for local features as well as global features. Two objective functions, namely, Mean Squared Error (MSE) and Peak Signal-to-Noise Ratio (PSNR), are employed for objective quality assessment between the estimated color image and its ground truth. The model is trained on the dataset created by ourselves with 1.2 K historical images comprised of old and ancient photographs of Nepal, each having 256 × 256 resolution. The loss i.e., MSE, PSNR, and accuracy of the model are found to be 6.08%, 34.65 dB, and 75.23%, respectively. Other than presenting the training results, the public acceptance or subjective validation of the generated images is assessed by means of a user study where the model shows 41.71% of naturalness while evaluating colorization results.

Download Full-text

Application of Rough and Fuzzy Set Theory for Prediction of Stochastic Wind Speed Data Using Long Short-Term Memory

Atmosphere ◽

10.3390/atmos12070924 ◽

2021 ◽

Vol 12 (7) ◽

pp. 924

Author(s):

Moslem Imani ◽

Hoda Fakour ◽

Wen-Hau Lan ◽

Huan-Chin Kao ◽

Chi Ming Lee ◽

...

Keyword(s):

Deep Learning ◽

Wind Speed ◽

Set Theory ◽

Rough Set ◽

Fuzzy Set Theory ◽

Short Term Memory ◽

Learning Model ◽

Short Term ◽

Long Short Term Memory ◽

Deep Learning Model

Despite the great significance of precisely forecasting the wind speed for development of the new and clean energy technology and stable grid operators, the stochasticity of wind speed makes the prediction a complex and challenging task. For improving the security and economic performance of power grids, accurate short-term wind power forecasting is crucial. In this paper, a deep learning model (Long Short-term Memory (LSTM)) has been proposed for wind speed prediction. Knowing that wind speed time series is nonlinear stochastic, the mutual information (MI) approach was used to find the best subset from the data by maximizing the joint MI between subset and target output. To enhance the accuracy and reduce input characteristics and data uncertainties, rough set and interval type-2 fuzzy set theory are combined in the proposed deep learning model. Wind speed data from an international airport station in the southern coast of Iran Bandar-Abbas City was used as the original input dataset for the optimized deep learning model. Based on the statistical results, the rough set LSTM (RST-LSTM) model showed better prediction accuracy than fuzzy and original LSTM, as well as traditional neural networks, with the lowest error for training and testing datasets in different time horizons. The suggested model can support the optimization of the control approach and the smooth procedure of power system. The results confirm the superior capabilities of deep learning techniques for wind speed forecasting, which could also inspire new applications in meteorology assessment.

Download Full-text

Tropical cyclone simulations over Bangladesh at convection permitting 4.4 km & 1.5 km resolution

Scientific Data ◽

10.1038/s41597-021-00847-5 ◽

2021 ◽

Vol 8 (1) ◽

Author(s):

Hamish Steptoe ◽

Nicholas Henry Savage ◽

Saeed Sadri ◽

Kate Salmon ◽

Zubair Maalick ◽

...

Keyword(s):

Wind Speed ◽

Tropical Cyclone ◽

Sea Level ◽

Tropical Cyclones ◽

Weather Forecasting ◽

Sea Level Pressure ◽

Mean Sea Level ◽

Level Pressure ◽

Medium Range ◽

Gust Speed

AbstractHigh resolution simulations at 4.4 km and 1.5 km resolution have been performed for 12 historical tropical cyclones impacting Bangladesh. We use the European Centre for Medium-Range Weather Forecasting 5th generation Re-Analysis (ERA5) to provide a 9-member ensemble of initial and boundary conditions for the regional configuration of the Met Office Unified Model. The simulations are compared to the original ERA5 data and the International Best Track Archive for Climate Stewardship (IBTrACS) tropical cyclone database for wind speed, gust speed and mean sea-level pressure. The 4.4 km simulations show a typical increase in peak gust speed of 41 to 118 knots relative to ERA5, and a deepening of minimum mean sea-level pressure of up to −27 hPa, relative to ERA5 and IBTrACS data. The downscaled simulations compare more favourably with IBTrACS data than the ERA5 data suggesting tropical cyclone hazards in the ERA5 deterministic output may be underestimated. The dataset is freely available from 10.5281/zenodo.3600201.

Download Full-text

Evaluation of NASA POWER Reanalysis Products to Estimate Daily Weather Variables in a Hot Summer Mediterranean Climate

Agronomy ◽

10.3390/agronomy11061207 ◽

2021 ◽

Vol 11 (6) ◽

pp. 1207

Author(s):

Gonçalo C. Rodrigues ◽

Ricardo P. Braga

Keyword(s):

Wind Speed ◽

Bias Correction ◽

Goodness Of Fit ◽

Mediterranean Climate ◽

Weather Data ◽

Coefficient Of Determination ◽

Mean Bias Error ◽

Bias Error ◽

Weather Variables ◽

Weather Stations

This study aims to evaluate NASA POWER reanalysis products for daily surface maximum (Tmax) and minimum (Tmin) temperatures, solar radiation (Rs), relative humidity (RH) and wind speed (Ws) when compared with observed data from 14 distributed weather stations across Alentejo Region, Southern Portugal, with a hot summer Mediterranean climate. Results showed that there is good agreement between NASA POWER reanalysis and observed data for all parameters, except for wind speed, with coefficient of determination (R2) higher than 0.82, with normalized root mean square error (NRMSE) varying, from 8 to 20%, and a normalized mean bias error (NMBE) ranging from –9 to 26%, for those variables. Based on these results, and in order to improve the accuracy of the NASA POWER dataset, two bias corrections were performed to all weather variables: one for the Alentejo Region as a whole; another, for each location individually. Results improved significantly, especially when a local bias correction is performed, with Tmax and Tmin presenting an improvement of the mean NRMSE of 6.6 °C (from 8.0 °C) and 16.1 °C (from 20.5 °C), respectively, while a mean NMBE decreased from 10.65 to 0.2%. Rs results also show a very high goodness of fit with a mean NRMSE of 11.2% and mean NMBE equal to 0.1%. Additionally, bias corrected RH data performed acceptably with an NRMSE lower than 12.1% and an NMBE below 2.1%. However, even when a bias correction is performed, Ws lacks the performance showed by the remaining weather variables, with an NRMSE never lower than 19.6%. Results show that NASA POWER can be useful for the generation of weather data sets where ground weather stations data is of missing or unavailable.

Download Full-text

Coupling between the JULES land-surface scheme and the CCATT-BRAMS atmospheric chemistry model (JULES-CCATT-BRAMS1.0): applications to numerical weather forecasting and the CO<sub>2</sub> budget in South America

Geoscientific Model Development Discussions ◽

10.5194/gmdd-6-453-2013 ◽

2013 ◽

Vol 6 (1) ◽

pp. 453-494 ◽

Cited By ~ 4

Author(s):

D. S. Moreira ◽

S. R. Freitas ◽

J. P. Bonatti ◽

L. M. Mercado ◽

N. M. É. Rosário ◽

...

Keyword(s):

South America ◽

Atmospheric Chemistry ◽

Land Surface ◽

Amazon Basin ◽

Mean Squared Error ◽

Weather Forecasting ◽

Dew Point ◽

Surface Model ◽

Co2 Concentrations ◽

Wide Range

Abstract. This article presents the development of a new numerical system denominated JULES-CCATT-BRAMS, which resulted from the coupling of the JULES surface model to the CCATT-BRAMS atmospheric chemistry model. The performance of this system in relation to several meteorological variables (wind speed at 10 m, air temperature at 2 m, dew point temperature at 2 m, pressure reduced to mean sea level and 6 h accumulated precipitation) and the CO2 concentration above an extensive area of South America is also presented, focusing on the Amazon basin. The evaluations were conducted for two periods, the wet (March) and dry (September) seasons of 2010. The statistics used to perform the evaluation included bias (BIAS) and root mean squared error (RMSE). The errors were calculated in relation to observations at conventional stations in airports and automatic stations. In addition, CO2 concentrations in the first model level were compared with meteorological tower measurements and vertical CO2 profiles were compared with aircraft data. The results of this study show that the JULES model coupled to CCATT-BRAMS provided a significant gain in performance in the evaluated atmospheric fields relative to those simulated by the LEAF (version 3) surface model originally utilized by CCATT-BRAMS. Simulations of CO2 concentrations in Amazonia and a comparison with observations are also discussed and show that the system presents a gain in performance relative to previous studies. Finally, we discuss a wide range of numerical studies integrating coupled atmospheric, land surface and chemistry processes that could be produced with the system described here. Therefore, this work presents to the scientific community a free tool, with good performance in relation to the observed data and re-analyses, able to produce atmospheric simulations/forecasts at different resolutions, for any period of time and in any region of the globe.

Download Full-text

Analysis on the Evolution and Microphysical Characteristics of Two Consecutive Hailstorms in Spring in Yunnan, China

Atmosphere ◽

10.3390/atmos12010063 ◽

2021 ◽

Vol 12 (1) ◽

pp. 63

Author(s):

Sidou Zhang ◽

Shiyin Liu ◽

Tengfei Zhang

Keyword(s):

Wind Speed ◽

Cloud Model ◽

Global Analysis ◽

Maximum Velocity ◽

Western Pacific Subtropical High ◽

Convective Cloud ◽

Radar Data ◽

Reanalysis Data ◽

Cloud Water ◽

High Level

By using products of the cloud model, National Centers for Environmental Prediction (NCEP) Final Operational Global Analysis (FNL) reanalysis data, and Doppler weather radar data, the mesoscale characteristics, microphysical structure, and mechanism of two hail cloud systems which occurred successively within 24 h in southeastern Yunnan have been analyzed. The results show that under the influence of two southwest jets in front of the south branch trough (SBT) and the periphery of the western Pacific subtropical high (WPSH), the northeast-southwest banded echoes affect the southeastern Yunnan of China twice. Meanwhile, the local mesoscale radial wind convergence and uneven wind speed lead to the intense development of convective echoes and the occurrence of hail. The simulated convective cloud bands are similar to the observation. The high-level mesoscale convergence line leads to the development of convective cloud bands. The low-level wind direction or wind speed convergence and the high-level wind speed divergence form a deep tilted updraft, with the maximum velocity of 15 m·s−1 at the −40~−10 °C layer, resulting in the intense development of local convective clouds. The hail embryos form through the conversion or collision growth of cloud water and snowflakes and have little to do with rain and ice crystals. Abundant cloud water, especially the accumulation region of high supercooled water (cloud water) near the 0 °C layer, is the key to the formation of hail embryos, in which qc is up to 1.92 g·kg−1 at the −4~−2 °C layer. The hail embryos mainly grow by collision-coalescence (collision-freezing) with cloud water (supercooled cloud drops) and snow crystal riming.

Download Full-text