Articles | Volume 9, issue 6
https://doi.org/10.5194/wes-9-1431-2024
© Author(s) 2024. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
https://doi.org/10.5194/wes-9-1431-2024
© Author(s) 2024. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Research article
| 
27 Jun 2024
Research article |
| 27 Jun 2024
| 27 Jun 2024
Machine-learning-based estimate of the wind speed over complex terrain using the long short-term memory (LSTM) recurrent neural network
Cássia Maria Leme Beu
CORRESPONDING AUTHOR
Instituto de Pesquisas Energéticas e Nucleares (IPEN), 2242 Prof. Lineu Prestes Av., São Paulo, Brazil
Eduardo Landulfo
Instituto de Pesquisas Energéticas e Nucleares (IPEN), 2242 Prof. Lineu Prestes Av., São Paulo, Brazil
Related authors
No articles found.
Hazel Vernier, Demilson Quintão, Bruno Biazon, Eduardo Landulfo, Giovanni Souza, V. Amanda Santos, J. S. Fabio Lopes, C. P. Alex Mendes, A. S. José da Matta, K. Pinheiro Damaris, Benoit Grosslin, P. M. P. Maria Jorge, Maria de Fátima Andrade, Neeraj Rastogi, Akhil Raj, Hongyu Liu, Mahesh Kovilakam, Suvarna Fadnavis, Frank G. Wienhold, Mathieu Colombier, D. Chris Boone, Gwenael Berthet, Nicolas Dumelie, Lilian Joly, and Jean-Paul Vernier
EGUsphere, https://doi.org/10.5194/egusphere-2025-924, https://doi.org/10.5194/egusphere-2025-924, 2025
This preprint is open for discussion and under review for Atmospheric Chemistry and Physics (ACP).
Short summary
Short summary
The eruption of Hunga Tonga-Hunga Ha'apai injected large amounts of water vapor and sea salt into the stratosphere, altering traditional views of volcanic aerosols. Using balloon-borne samplers, we collected aerosol samples and found high levels of sea salt and calcium, suggesting sulfate depletion due to gypsum formation. These findings highlight the need to consider sea salt in climate models to better predict volcanic impacts on the atmosphere and climate.
Juan Vicente Pallotta, Silvânia Alves de Carvalho, Fabio Juliano da Silva Lopes, Alexandre Cacheffo, Eduardo Landulfo, and Henrique Melo Jorge Barbosa
Geosci. Instrum. Method. Data Syst., 12, 171–185, https://doi.org/10.5194/gi-12-171-2023, https://doi.org/10.5194/gi-12-171-2023, 2023
Short summary
Short summary
Lidar networks coordinate efforts of different groups, providing guidelines to homogenize retrievals from different instruments. We describe an ongoing effort to develop the Lidar Processing Pipeline (LPP) collaboratively, a collection of tools developed in C/C++ to handle all the steps of a typical lidar analysis. Analysis of simulations and real lidar data showcases the LPP’s features. From this exercise, we draw a roadmap to guide future development, accommodating the needs of our community.
Gregori de Arruda Moreira, Fábio Juliano da Silva Lopes, Juan Luis Guerrero-Rascado, Jonatan João da Silva, Antonio Arleques Gomes, Eduardo Landulfo, and Lucas Alados-Arboledas
Atmos. Meas. Tech., 12, 4261–4276, https://doi.org/10.5194/amt-12-4261-2019, https://doi.org/10.5194/amt-12-4261-2019, 2019
Short summary
Short summary
In this paper, we present a comparative analysis of the use of lidar-backscattered signals at three wavelengths (355, 532 and 1064 nm) to study the ABL by investigating high-order moments, which gives us information about the ABL height (derived using the variance method), aerosol layer movements (skewness) and mixing conditions (kurtosis) at several heights.
Gregori de Arruda Moreira, Juan Luis Guerrero-Rascado, Jose A. Benavent-Oltra, Pablo Ortiz-Amezcua, Roberto Román, Andrés E. Bedoya-Velásquez, Juan Antonio Bravo-Aranda, Francisco Jose Olmo Reyes, Eduardo Landulfo, and Lucas Alados-Arboledas
Atmos. Chem. Phys., 19, 1263–1280, https://doi.org/10.5194/acp-19-1263-2019, https://doi.org/10.5194/acp-19-1263-2019, 2019
Short summary
Short summary
In this study we show the capabilities of combining different remote sensing systems (microwave radiometer – MWR, Doppler lidar – DL – and elastic lidar – EL) for retrieving a detailed picture of the PBL turbulent features. Concerning EL, in addition to analyzing the influence of noise, we explore the use of different wavelengths, which usually includes EL systems operated in extended networks, like EARLINET, LALINET, MPLNET or SKYNET.
Igor Veselovskii, Philippe Goloub, Qiaoyun Hu, Thierry Podvin, David N. Whiteman, Mikhael Korenskiy, and Eduardo Landulfo
Atmos. Meas. Tech., 12, 119–128, https://doi.org/10.5194/amt-12-119-2019, https://doi.org/10.5194/amt-12-119-2019, 2019
Short summary
Short summary
Methane is currently the second most important greenhouse gas of anthropogenic origin (after carbon dioxide) and its concentration can be increased inside the boundary layer. So, the development of instruments for vertical profiling of the methane mixing ratio is an important task. We present the results of methane profiling in the lower troposphere using LILAS Raman lidar from the Lille University observatory platform (France).
Carlos Eduardo Souto-Oliveira, Maria de Fátima Andrade, Prashant Kumar, Fábio Juliano da Silva Lopes, Marly Babinski, and Eduardo Landulfo
Atmos. Chem. Phys., 16, 14635–14656, https://doi.org/10.5194/acp-16-14635-2016, https://doi.org/10.5194/acp-16-14635-2016, 2016
Short summary
Short summary
The Metropolitan Area of São Paulo is the biggest megacity of South America, with over 20 million inhabitants. In recent years, the region has been facing a modification in rain patterns. In this study, we evaluated the effects of local and remote sources of air pollution on cloud-condensation nuclei activation properties. Our results showed that the local vehicular traffic emission products presented more negative effects on cloud-condensation nuclei activation than the remote sources.
F. J. S. Lopes, E. Landulfo, and M. A. Vaughan
Atmos. Meas. Tech., 6, 3281–3299, https://doi.org/10.5194/amt-6-3281-2013, https://doi.org/10.5194/amt-6-3281-2013, 2013
E. G. Larroza, W. M. Nakaema, R. Bourayou, C. Hoareau, E. Landulfo, and P. Keckhut
Atmos. Meas. Tech., 6, 3197–3210, https://doi.org/10.5194/amt-6-3197-2013, https://doi.org/10.5194/amt-6-3197-2013, 2013
Related subject area
Thematic area: Wind and the atmosphere | Topic: Wind and turbulence
Flow acceleration statistics: a new paradigm for wind-driven loads, towards probabilistic turbine design
Observations of wind farm wake recovery at an operating wind farm
Periods of constant wind speed: how long do they last in the turbulent atmospheric boundary layer?
Characterization of local wind profiles: a random forest approach for enhanced wind profile extrapolation
Simulations suggest offshore wind farms modify low-level jets
On the lidar-turbulence paradox and possible countermeasures
The actuator farm model for large eddy simulation (LES) of wind-farm-induced atmospheric gravity waves and farm–farm interaction
Understanding the impact of data gaps on long-term offshore wind resource estimates
Evaluating mesoscale model predictions of diurnal speedup events in the Altamont Pass Wind Resource Area of California
Converging profile relationships for offshore wind speed and turbulence intensity
An analytical formulation for turbulent kinetic energy added by wind turbines based on large-eddy simulation
A simple steady-state inflow model of the neutral and stable atmospheric boundary layer applied to wind turbine wake simulations
Influences of lidar scanning parameters on wind turbine wake retrievals in complex terrain
Tall Wind Profile Validation Using Lidar Observations and Hindcast Data
Experimental evaluation of wind turbine wake turbulence impacts on a general aviation aircraft
Underestimation of strong wind speeds offshore in ERA5: evidence, discussion and correction
Brief communication: A simple axial induction modification to the Weather Research and Forecasting Fitch wind farm parameterization
Impact of swell waves on atmospheric surface turbulence: wave–turbulence decomposition methods
Method to predict the minimum measurement and experiment durations needed to achieve converged and significant results in a wind energy field experiment
Meteorological Impacts of Offshore Wind Turbines as Simulated in the Weather Research and Forecasting Model
Swell Impacts on an Offshore Wind Farm in Stable Boundary Layer: Wake Flow and Energy Budget Analysis
Evaluation of wind farm parameterizations in the WRF model under different atmospheric stability conditions with high-resolution wake simulations
Renewable Energy Complementarity (RECom) maps – a comprehensive visualisation tool to support spatial diversification
Control-oriented modelling of wind direction variability
Machine learning methods to improve spatial predictions of coastal wind speed profiles and low-level jets using single-level ERA5 data
Offshore low-level jet observations and model representation using lidar buoy data off the California coast
Measurement-driven large-eddy simulations of a diurnal cycle during a wake-steering field campaign
The fractal turbulent–non-turbulent interface in the atmosphere
TOSCA – an open-source, finite-volume, large-eddy simulation (LES) environment for wind farm flows
Quantitative comparison of power production and power quality onshore and offshore: a case study from the eastern United States
The wind farm pressure field
Realistic turbulent inflow conditions for estimating the performance of a floating wind turbine
Brief communication: On the definition of the low-level jet
A decision-tree-based measure–correlate–predict approach for peak wind gust estimation from a global reanalysis dataset
Revealing inflow and wake conditions of a 6 MW floating turbine
Stochastic gradient descent for wind farm optimization
Modelling the impact of trapped lee waves on offshore wind farm power output
Applying a random time mapping to Mann-modeled turbulence for the generation of intermittent wind fields
From shear to veer: theory, statistics, and practical application
Quantification and correction of motion influence for nacelle-based lidar systems on floating wind turbines
Gaussian mixture models for the optimal sparse sampling of offshore wind resource
Dependence of turbulence estimations on nacelle lidar scanning strategies
Vertical extrapolation of Advanced Scatterometer (ASCAT) ocean surface winds using machine-learning techniques
An investigation of spatial wind direction variability and its consideration in engineering models
From gigawatt to multi-gigawatt wind farms: wake effects, energy budgets and inertial gravity waves investigated by large-eddy simulations
Investigations of correlation and coherence in turbulence from a large-eddy simulation
Validation of turbulence intensity as simulated by the Weather Research and Forecasting model off the US northeast coast
On the laminar–turbulent transition mechanism on megawatt wind turbine blades operating in atmospheric flow
Brief communication: A momentum-conserving superposition method applied to the super-Gaussian wind turbine wake model
Turbulence structures and entrainment length scales in large offshore wind farms
Mark Kelly
Wind Energ. Sci., 10, 535–558, https://doi.org/10.5194/wes-10-535-2025, https://doi.org/10.5194/wes-10-535-2025, 2025
Short summary
Short summary
Industrial standards for wind turbine design are based on 10 min statistics of wind speed at turbine hub height, treating fluctuations as turbulence. But recent work shows that the effect of strong transients is described by flow acceleration. We devise a method to measure the acceleration that turbines encounter; the extremes offshore defy 10 min averages due to various phenomena beyond turbulence. These findings are translated into a recipe supporting statistical design.
Raghavendra Krishnamurthy, Rob K. Newsom, Colleen M. Kaul, Stefano Letizia, Mikhail Pekour, Nicholas Hamilton, Duli Chand, Donna Flynn, Nicola Bodini, and Patrick Moriarty
Wind Energ. Sci., 10, 361–380, https://doi.org/10.5194/wes-10-361-2025, https://doi.org/10.5194/wes-10-361-2025, 2025
Short summary
Short summary
This study examines how atmospheric phenomena affect the recovery of wind farm wake – the disturbed air behind turbines. In regions like Oklahoma, where wind farms are often clustered, understanding wake recovery is crucial. We found that wind farms can alter phenomena like low-level jets, which are common in Oklahoma, by deflecting them above the wind farm. As a result, the impact of wakes can be observed up to 1–2 km above ground level.
Daniela Moreno, Jan Friedrich, Matthias Wächter, Jörg Schwarte, and Joachim Peinke
Wind Energ. Sci., 10, 347–360, https://doi.org/10.5194/wes-10-347-2025, https://doi.org/10.5194/wes-10-347-2025, 2025
Short summary
Short summary
Unexpected load events measured on operating wind turbines are not accurately predicted by numerical simulations. We introduce the periods of constant wind speed as a possible cause of such events. We measure and characterize their statistics from atmospheric data. Further comparisons to standard modelled data and experimental turbulence data suggest that such events are not intrinsic to small-scale turbulence and are not accurately described by current standard wind models.
Farkhondeh (Hanie) Rouholahnejad and Julia Gottschall
Wind Energ. Sci., 10, 143–159, https://doi.org/10.5194/wes-10-143-2025, https://doi.org/10.5194/wes-10-143-2025, 2025
Short summary
Short summary
In wind energy, precise wind speed prediction at hub height is vital. Our study in the Dutch North Sea reveals that the on-site-trained random forest model outperforms the global reanalysis data, ERA5, in accuracy and precision. Trained within a 200 km range, the model effectively extends the wind speed vertically but experiences bias. It also outperforms ERA5 corrected with measurements in capturing wind speed variations and fine wind patterns, highlighting its potential for site assessment.
Daphne Quint, Julie K. Lundquist, and David Rosencrans
Wind Energ. Sci., 10, 117–142, https://doi.org/10.5194/wes-10-117-2025, https://doi.org/10.5194/wes-10-117-2025, 2025
Short summary
Short summary
Offshore wind farms will be built along the East Coast of the United States. Low-level jets (LLJs) – layers of fast winds at low altitudes – also occur here. LLJs provide wind resources and also influence moisture and pollution transport, so it is important to understand how they might change. We develop and validate an automated tool to detect LLJs and compare 1 year of simulations with and without wind farms. Here, we describe LLJ characteristics and how they change with wind farms.
Alfredo Peña, Ginka G. Yankova, and Vasiliki Mallini
Wind Energ. Sci., 10, 83–102, https://doi.org/10.5194/wes-10-83-2025, https://doi.org/10.5194/wes-10-83-2025, 2025
Short summary
Short summary
Lidars are vastly used in wind energy, but most users struggle when interpreting lidar turbulence measures. Here, we explain the difficulty in converting them into standard measurements. We show two ways of converting lidar to in situ turbulence measurements, both using neural networks: one of them is based on physics, while the other is purely data-driven. They show promising results when compared to high-quality turbulence measurements from a tall mast.
Sebastiano Stipa, Arjun Ajay, and Joshua Brinkerhoff
Wind Energ. Sci., 9, 2301–2332, https://doi.org/10.5194/wes-9-2301-2024, https://doi.org/10.5194/wes-9-2301-2024, 2024
Short summary
Short summary
This study presents the actuator farm model, a new method for modeling wind turbines within large wind farms. The model greatly reduces computational cost when compared to traditional actuator wind turbine models and is beneficial for studying flow around large wind farms as well as the interaction between multiple wind farms. Results obtained from numerical simulations show excellent agreement with past wind turbine models, demonstrating its utility for future large-scale wind farm simulations.
Martin Georg Jonietz Alvarez, Warren Watson, and Julia Gottschall
Wind Energ. Sci., 9, 2217–2233, https://doi.org/10.5194/wes-9-2217-2024, https://doi.org/10.5194/wes-9-2217-2024, 2024
Short summary
Short summary
Offshore wind measurements are often affected by gaps. We investigated how these gaps affect wind resource assessments and whether filling them reduces their effect. We find that the effect of gaps on the estimated long-term wind resource is lower than expected and that data gap filling does not significantly change the outcome. These results indicate a need to reduce current wind data availability requirements for offshore measurement campaigns.
Robert S. Arthur, Alex Rybchuk, Timothy W. Juliano, Gabriel Rios, Sonia Wharton, Julie K. Lundquist, and Jerome D. Fast
Wind Energ. Sci. Discuss., https://doi.org/10.5194/wes-2024-137, https://doi.org/10.5194/wes-2024-137, 2024
Revised manuscript accepted for WES
Short summary
Short summary
This paper evaluates a new model configuration for wind energy forecasting in complex terrain. We compare model results to observations in the Altamont Pass (California, USA), where wind channeling through a mountain pass leads to increased energy production. We show evidence of improved wind speed and turbulence predictions compared to a more established modeling approach. Our work helps to ensure the robustness of the new model configuration for future wind energy applications.
Gus Jeans
Wind Energ. Sci., 9, 2001–2015, https://doi.org/10.5194/wes-9-2001-2024, https://doi.org/10.5194/wes-9-2001-2024, 2024
Short summary
Short summary
An extensive set of met mast data offshore northwestern Europe are used to reduce uncertainty in offshore wind speed and turbulence intensity. The performance of widely used industry standard relationships is quantified, while some new empirical relationships are derived for practical application. Motivations include encouraging appropriate convergence of traditionally separate technical disciplines within the rapidly growing offshore wind energy industry.
Ali Khanjari, Asim Feroz, and Cristina Archer
Wind Energ. Sci. Discuss., https://doi.org/10.5194/wes-2024-128, https://doi.org/10.5194/wes-2024-128, 2024
Revised manuscript accepted for WES
Short summary
Short summary
Wind turbines add turbulence to the atmosphere behind them, especially 4–6 diameters downstream and near the rotor top. We propose an equation that predicts the distribution of added turbulence as a function of a turbine parameter (thrust coefficient) and an atmospheric parameter (undisturbed turbulence intensity before the turbine). We find that our equation performs well, although not perfectly. Eventually this equation can be used to better understand how wind turbines affect the atmosphere.
Maarten Paul van der Laan, Mark Kelly, Mads Baungaard, Antariksh Dicholkar, and Emily Louise Hodgson
Wind Energ. Sci., 9, 1985–2000, https://doi.org/10.5194/wes-9-1985-2024, https://doi.org/10.5194/wes-9-1985-2024, 2024
Short summary
Short summary
Wind turbines are increasing in size and operate more frequently above the atmospheric surface layer, which requires improved inflow models for numerical simulations of turbine interaction. In this work, a novel steady-state model of the atmospheric boundary layer (ABL) is introduced. Numerical wind turbine flow simulations subjected to shallow and tall ABLs are conducted, and the proposed model shows improved performance compared to other state-of-the-art steady-state models.
Rachel Robey and Julie K. Lundquist
Wind Energ. Sci., 9, 1905–1922, https://doi.org/10.5194/wes-9-1905-2024, https://doi.org/10.5194/wes-9-1905-2024, 2024
Short summary
Short summary
Measurements of wind turbine wakes with scanning lidar instruments contain complex errors. We model lidars in a simulated environment to understand how and why the measured wake may differ from the true wake and validate the results with observational data. The lidar smooths out the wake, making it seem more spread out and the slowdown of the winds less pronounced. Our findings provide insights into best practices for accurately measuring wakes with lidar and interpreting observational data.
Etienne Cheynet, Jan Markus Diezel, Hilde Haakenstad, Øyvind Breivik, Alfredo Peña, and Joachim Reuder
Wind Energ. Sci. Discuss., https://doi.org/10.5194/wes-2024-119, https://doi.org/10.5194/wes-2024-119, 2024
Revised manuscript accepted for WES
Short summary
Short summary
This study aims to help future large offshore wind turbines and airborne wind energy systems by providing insights into wind speeds at much higher altitudes than previously examined. We assessed three wind models (ERA5, NORA3, and NEWA) to predict wind speeds up to 500 m. Using lidar data from Norway and the North Sea, we found that ERA5 excels offshore, while NORA3 performs best onshore. However, the performance of the models depends on the locations and the evaluation criteria.
Jonathan D. Rogers
Wind Energ. Sci., 9, 1849–1868, https://doi.org/10.5194/wes-9-1849-2024, https://doi.org/10.5194/wes-9-1849-2024, 2024
Short summary
Short summary
This paper describes the results of a flight experiment to assess the existence of potential safety risks to a general aviation aircraft from added turbulence in the wake of a wind turbine. A general aviation aircraft was flown through the wake of an operating wind turbine at different downwind distances. Results indicated that there were small increases in disturbances to the aircraft due to added turbulence in the wake, but they never approached levels that would pose a safety risk.
Rémi Gandoin and Jorge Garza
Wind Energ. Sci., 9, 1727–1745, https://doi.org/10.5194/wes-9-1727-2024, https://doi.org/10.5194/wes-9-1727-2024, 2024
Short summary
Short summary
ERA5 has become the workhorse of most wind resource assessment applications, as it compares better with in situ measurements than other reanalyses. However, for design purposes, ERA5 suffers from a drawback: it underestimates strong wind speeds offshore (approx. from 10 m s−1). This is not widely discussed in the scientific literature. We address this bias and proposes a simple, robust correction. This article supports the growing need for use-case-specific validations of reanalysis datasets.
Lukas Vollmer, Balthazar Arnoldus Maria Sengers, and Martin Dörenkämper
Wind Energ. Sci., 9, 1689–1693, https://doi.org/10.5194/wes-9-1689-2024, https://doi.org/10.5194/wes-9-1689-2024, 2024
Short summary
Short summary
This study proposes a modification to a well-established wind farm parameterization used in mesoscale models. The wind speed at the location of the turbine, which is used to calculate power and thrust, is corrected to approximate the free wind speed. Results show that the modified parameterization produces more accurate estimates of the turbine’s power curve.
Mostafa Bakhoday Paskyabi
Wind Energ. Sci., 9, 1631–1645, https://doi.org/10.5194/wes-9-1631-2024, https://doi.org/10.5194/wes-9-1631-2024, 2024
Short summary
Short summary
The exchange of momentum and energy between the atmosphere and ocean depends on air–sea processes, especially wave-related ones. Precision in representing these interactions is vital for offshore wind turbine and farm design and operation. The development of a reliable wave–turbulence decomposition method to remove wave-induced interference from single-height wind measurements is essential for these applications and enhances our grasp of wind coherence within the wave boundary layer.
Daniel R. Houck, Nathaniel B. de Velder, David C. Maniaci, and Brent C. Houchens
Wind Energ. Sci., 9, 1189–1209, https://doi.org/10.5194/wes-9-1189-2024, https://doi.org/10.5194/wes-9-1189-2024, 2024
Short summary
Short summary
Experiments offer incredible value to science, but results must come with an uncertainty quantification to be meaningful. We present a method to simulate a proposed experiment, calculate uncertainties, and determine the measurement duration (total time of measurements) and the experiment duration (total time to collect the required measurement data when including condition variability and time when measurement is not occurring) required to produce statistically significant and converged results.
Daphne Quint, Julie K. Lundquist, Nicola Bodini, and David Rosencrans
Wind Energ. Sci. Discuss., https://doi.org/10.5194/wes-2024-53, https://doi.org/10.5194/wes-2024-53, 2024
Revised manuscript accepted for WES
Short summary
Short summary
Offshore wind farms along the US east coast can have limited effects on local weather. Studying this, we used a weather model to compare conditions with and without wind farms near Massachusetts and Rhode Island. We analyzed changes in wind, temperature, and turbulence. Results show reduced wind speeds near and downwind of wind farms, especially during stability and high winds. Turbulence increases near wind farms, affecting boundary-layer height and wake size.
Xu Ning and Mostafa Bakhoday-Paskyabi
Wind Energ. Sci. Discuss., https://doi.org/10.5194/wes-2024-38, https://doi.org/10.5194/wes-2024-38, 2024
Revised manuscript accepted for WES
Short summary
Short summary
Waves interact with the overlying wind field by modifying the stresses at the atmosphere-ocean interface. We develop and employ a parameterization method of wave-induced stresses in the numerical simulation of an offshore wind farm in a stable atmospheric boundary layer. This work demonstrates how swells change the kinetic energy transport, and induce wind veer and wake deflection, leading to significant variations in the power output of wind turbines at different positions of the wind farm.
Oscar García-Santiago, Andrea N. Hahmann, Jake Badger, and Alfredo Peña
Wind Energ. Sci., 9, 963–979, https://doi.org/10.5194/wes-9-963-2024, https://doi.org/10.5194/wes-9-963-2024, 2024
Short summary
Short summary
This study compares the results of two wind farm parameterizations (WFPs) in the Weather Research and Forecasting model, simulating a two-turbine array under three atmospheric stabilities with large-eddy simulations. We show that the WFPs accurately depict wind speeds either near turbines or in the far-wake areas, but not both. The parameterizations’ performance varies by variable (wind speed or turbulent kinetic energy) and atmospheric stability, with reduced accuracy in stable conditions.
Til Kristian Vrana and Harald G. Svendsen
Wind Energ. Sci., 9, 919–932, https://doi.org/10.5194/wes-9-919-2024, https://doi.org/10.5194/wes-9-919-2024, 2024
Short summary
Short summary
We developed new ways to plot comprehensive wind resource maps that show the revenue potential of different locations for future wind power developments. The relative capacity factor is introduced as an indicator showing the expected mean power output. The market value factor is introduced, which captures the expected mean market value relative to other wind parks. The Renewable Energy Complementarity (RECom) index combines the two into a single index, resulting in the RECom map.
Scott Dallas, Adam Stock, and Edward Hart
Wind Energ. Sci., 9, 841–867, https://doi.org/10.5194/wes-9-841-2024, https://doi.org/10.5194/wes-9-841-2024, 2024
Short summary
Short summary
This review presents the current understanding of wind direction variability in the context of control-oriented modelling of wind turbines and wind farms in a manner suitable to a wide audience. Motivation comes from the significant and commonly seen yaw error of horizontal axis wind turbines, which carries substantial negative impacts on annual energy production and the levellised cost of wind energy. Gaps in the literature are identified, and the critical challenges in this area are discussed.
Christoffer Hallgren, Jeanie A. Aird, Stefan Ivanell, Heiner Körnich, Ville Vakkari, Rebecca J. Barthelmie, Sara C. Pryor, and Erik Sahlée
Wind Energ. Sci., 9, 821–840, https://doi.org/10.5194/wes-9-821-2024, https://doi.org/10.5194/wes-9-821-2024, 2024
Short summary
Short summary
Knowing the wind speed across the rotor of a wind turbine is key in making good predictions of the power production. However, models struggle to capture both the speed and the shape of the wind profile. Using machine learning methods based on the model data, we show that the predictions can be improved drastically. The work focuses on three coastal sites, spread over the Northern Hemisphere (the Baltic Sea, the North Sea, and the US Atlantic coast) with similar results for all sites.
Lindsay M. Sheridan, Raghavendra Krishnamurthy, William I. Gustafson Jr., Ye Liu, Brian J. Gaudet, Nicola Bodini, Rob K. Newsom, and Mikhail Pekour
Wind Energ. Sci., 9, 741–758, https://doi.org/10.5194/wes-9-741-2024, https://doi.org/10.5194/wes-9-741-2024, 2024
Short summary
Short summary
In 2020, lidar-mounted buoys owned by the US Department of Energy (DOE) were deployed off the California coast in two wind energy lease areas and provided valuable year-long analyses of offshore low-level jet (LLJ) characteristics at heights relevant to wind turbines. In addition to the LLJ climatology, this work provides validation of LLJ representation in atmospheric models that are essential for assessing the potential energy yield of offshore wind farms.
Eliot Quon
Wind Energ. Sci., 9, 495–518, https://doi.org/10.5194/wes-9-495-2024, https://doi.org/10.5194/wes-9-495-2024, 2024
Short summary
Short summary
Engineering models used to design wind farms generally do not account for realistic atmospheric conditions that can rapidly evolve from minute to minute. This paper uses a first-principles simulation technique to predict the performance of five wind turbines during a wind farm control experiment. Challenges included limited observations and atypical conditions. The simulation accurately predicts the aerodynamics of a turbine when it is situated partially within the wake of an upstream turbine.
Lars Neuhaus, Matthias Wächter, and Joachim Peinke
Wind Energ. Sci., 9, 439–452, https://doi.org/10.5194/wes-9-439-2024, https://doi.org/10.5194/wes-9-439-2024, 2024
Short summary
Short summary
Future wind turbines reach unprecedented heights and are affected by wind conditions that have not yet been studied in detail. With increasing height, a transition to laminar conditions with a turbulent–non-turbulent interface (TNTI) becomes more likely. In this paper, the presence and fractality of this TNTI in the atmosphere are studied. Typical fractalities known from ideal laboratory and numerical studies and a frequent occurrence of the TNTI at heights of multi-megawatt turbines are found.
Sebastiano Stipa, Arjun Ajay, Dries Allaerts, and Joshua Brinkerhoff
Wind Energ. Sci., 9, 297–320, https://doi.org/10.5194/wes-9-297-2024, https://doi.org/10.5194/wes-9-297-2024, 2024
Short summary
Short summary
In the current study, we introduce TOSCA (Toolbox fOr Stratified Convective Atmospheres), an open-source computational fluid dynamics (CFD) tool, and demonstrate its capabilities by simulating the flow around a large wind farm, operating in realistic flow conditions. This is one of the grand challenges of the present decade and can yield better insight into physical phenomena that strongly affect wind farm operation but which are not yet fully understood.
Rebecca Foody, Jacob Coburn, Jeanie A. Aird, Rebecca J. Barthelmie, and Sara C. Pryor
Wind Energ. Sci., 9, 263–280, https://doi.org/10.5194/wes-9-263-2024, https://doi.org/10.5194/wes-9-263-2024, 2024
Short summary
Short summary
Using lidar-derived wind speed measurements at approx. 150 m height at onshore and offshore locations, we quantify the advantages of deploying wind turbines offshore in terms of the amount of electrical power produced and the higher reliability and predictability of the electrical power.
Ronald B. Smith
Wind Energ. Sci., 9, 253–261, https://doi.org/10.5194/wes-9-253-2024, https://doi.org/10.5194/wes-9-253-2024, 2024
Short summary
Short summary
Recent papers have investigated the impact of turbine drag on local wind patterns, but these studies have not given a full explanation of the induced pressure field. The pressure field blocks and deflects the wind and in other ways modifies farm efficiency. Current gravity wave models are complex and provide no estimation tools. We dig deeper into the cause of the pressure field and provide approximate closed-form expressions for pressure field effects.
Cédric Raibaudo, Jean-Christophe Gilloteaux, and Laurent Perret
Wind Energ. Sci., 8, 1711–1725, https://doi.org/10.5194/wes-8-1711-2023, https://doi.org/10.5194/wes-8-1711-2023, 2023
Short summary
Short summary
The work presented here proposes interfacing experimental measurements performed in a wind tunnel with simulations conducted with the aeroelastic code FAST and applied to a floating wind turbine model under wave-induced motion. FAST simulations using experiments match well with those obtained using the inflow generation method provided by TurbSim. The highest surge motion frequencies show a significant decrease in the mean power produced by the turbine and a mitigation of the flow dynamics.
Christoffer Hallgren, Jeanie A. Aird, Stefan Ivanell, Heiner Körnich, Rebecca J. Barthelmie, Sara C. Pryor, and Erik Sahlée
Wind Energ. Sci., 8, 1651–1658, https://doi.org/10.5194/wes-8-1651-2023, https://doi.org/10.5194/wes-8-1651-2023, 2023
Short summary
Short summary
Low-level jets (LLJs) are special types of non-ideal wind profiles affecting both wind energy production and loads on a wind turbine. However, among LLJ researchers, there is no consensus regarding which definition to use to identify these profiles. In this work, we compare two different ways of identifying the LLJ – the falloff definition and the shear definition – and argue why the shear definition is better suited to wind energy applications.
Serkan Kartal, Sukanta Basu, and Simon J. Watson
Wind Energ. Sci., 8, 1533–1551, https://doi.org/10.5194/wes-8-1533-2023, https://doi.org/10.5194/wes-8-1533-2023, 2023
Short summary
Short summary
Peak wind gust is a crucial meteorological variable for wind farm planning and operations. Unfortunately, many wind farms do not have on-site measurements of it. In this paper, we propose a machine-learning approach (called INTRIGUE, decIsioN-TRee-based wInd GUst Estimation) that utilizes numerous inputs from a public-domain reanalysis dataset, generating long-term, site-specific peak wind gust series.
Nikolas Angelou, Jakob Mann, and Camille Dubreuil-Boisclair
Wind Energ. Sci., 8, 1511–1531, https://doi.org/10.5194/wes-8-1511-2023, https://doi.org/10.5194/wes-8-1511-2023, 2023
Short summary
Short summary
This study presents the first experimental investigation using two nacelle-mounted wind lidars that reveal the upwind and downwind conditions relative to a full-scale floating wind turbine. We find that in the case of floating wind turbines with small pitch and roll oscillating motions (< 1°), the ambient turbulence is the main driving factor that determines the propagation of the wake characteristics.
Julian Quick, Pierre-Elouan Rethore, Mads Mølgaard Pedersen, Rafael Valotta Rodrigues, and Mikkel Friis-Møller
Wind Energ. Sci., 8, 1235–1250, https://doi.org/10.5194/wes-8-1235-2023, https://doi.org/10.5194/wes-8-1235-2023, 2023
Short summary
Short summary
Wind turbine positions are often optimized to avoid wake losses. These losses depend on atmospheric conditions, such as the wind speed and direction. The typical optimization scheme involves discretizing the atmospheric inputs, then considering every possible set of these discretized inputs in every optimization iteration. This work presents stochastic gradient descent (SGD) as an alternative, which randomly samples the atmospheric conditions during every optimization iteration.
Sarah J. Ollier and Simon J. Watson
Wind Energ. Sci., 8, 1179–1200, https://doi.org/10.5194/wes-8-1179-2023, https://doi.org/10.5194/wes-8-1179-2023, 2023
Short summary
Short summary
This modelling study shows that topographic trapped lee waves (TLWs) modify flow behaviour and power output in offshore wind farms. We demonstrate that TLWs can substantially alter the wind speeds at individual wind turbines and effect the power output of the turbine and whole wind farm. The impact on wind speeds and power is dependent on which part of the TLW wave cycle interacts with the wind turbines and wind farm. Positive and negative impacts of TLWs on power output are observed.
Khaled Yassin, Arne Helms, Daniela Moreno, Hassan Kassem, Leo Höning, and Laura J. Lukassen
Wind Energ. Sci., 8, 1133–1152, https://doi.org/10.5194/wes-8-1133-2023, https://doi.org/10.5194/wes-8-1133-2023, 2023
Short summary
Short summary
The current turbulent wind field models stated in the IEC 61400-1 standard underestimate the probability of extreme changes in wind velocity. This underestimation can lead to the false calculation of extreme and fatigue loads on the turbine. In this work, we are trying to apply a random time-mapping technique to one of the standard turbulence models to adapt to such extreme changes. The turbulent fields generated are compared with a standard wind field to show the effects of this new mapping.
Mark Kelly and Maarten Paul van der Laan
Wind Energ. Sci., 8, 975–998, https://doi.org/10.5194/wes-8-975-2023, https://doi.org/10.5194/wes-8-975-2023, 2023
Short summary
Short summary
The turning of the wind with height, which is known as veer, can affect wind turbine performance. Thus far meteorology has only given idealized descriptions of veer, which has not yet been related in a way readily usable for wind energy. Here we derive equations for veer in terms of meteorological quantities and provide practically usable forms in terms of measurable shear (change in wind speed with height). Flow simulations and measurements at turbine heights support these developments.
Moritz Gräfe, Vasilis Pettas, Julia Gottschall, and Po Wen Cheng
Wind Energ. Sci., 8, 925–946, https://doi.org/10.5194/wes-8-925-2023, https://doi.org/10.5194/wes-8-925-2023, 2023
Short summary
Short summary
Inflow wind field measurements from nacelle-based lidar systems offer great potential for different applications including turbine control, load validation and power performance measurements. On floating wind turbines nacelle-based lidar measurements are affected by the dynamic behavior of the floating foundations. Therefore, the effects on lidar wind speed measurements induced by floater dynamics must be well understood. A new model for quantification of these effects is introduced in our work.
Robin Marcille, Maxime Thiébaut, Pierre Tandeo, and Jean-François Filipot
Wind Energ. Sci., 8, 771–786, https://doi.org/10.5194/wes-8-771-2023, https://doi.org/10.5194/wes-8-771-2023, 2023
Short summary
Short summary
A novel data-driven method is proposed to design an optimal sensor network for the reconstruction of offshore wind resources. Based on unsupervised learning of numerical weather prediction wind data, it provides a simple yet efficient method for the siting of sensors, outperforming state-of-the-art methods for this application. It is applied in the main French offshore wind energy development areas to provide guidelines for the deployment of floating lidars for wind resource assessment.
Wei Fu, Alessandro Sebastiani, Alfredo Peña, and Jakob Mann
Wind Energ. Sci., 8, 677–690, https://doi.org/10.5194/wes-8-677-2023, https://doi.org/10.5194/wes-8-677-2023, 2023
Short summary
Short summary
Nacelle lidars with different beam scanning locations and two types of systems are considered for inflow turbulence estimations using both numerical simulations and field measurements. The turbulence estimates from a sonic anemometer at the hub height of a Vestas V52 turbine are used as references. The turbulence parameters are retrieved using the radial variances and a least-squares procedure. The findings from numerical simulations have been verified by the analysis of the field measurements.
Daniel Hatfield, Charlotte Bay Hasager, and Ioanna Karagali
Wind Energ. Sci., 8, 621–637, https://doi.org/10.5194/wes-8-621-2023, https://doi.org/10.5194/wes-8-621-2023, 2023
Short summary
Short summary
Wind observations at heights relevant to the operation of modern offshore wind farms, i.e. 100 m and more, are required to optimize their positioning and layout. Satellite wind retrievals provide observations of the wind field over large spatial areas and extensive time periods, yet their temporal resolution is limited and they are only representative at 10 m height. Machine-learning models are applied to lift these satellite winds to higher heights, directly relevant to wind energy purposes.
Anna von Brandis, Gabriele Centurelli, Jonas Schmidt, Lukas Vollmer, Bughsin' Djath, and Martin Dörenkämper
Wind Energ. Sci., 8, 589–606, https://doi.org/10.5194/wes-8-589-2023, https://doi.org/10.5194/wes-8-589-2023, 2023
Short summary
Short summary
We propose that considering large-scale wind direction changes in the computation of wind farm cluster wakes is of high relevance. Consequently, we present a new solution for engineering modeling tools that accounts for the effect of such changes in the propagation of wakes. The new model is evaluated with satellite data in the German Bight area. It has the potential to reduce uncertainty in applications such as site assessment and short-term power forecasting.
Oliver Maas
Wind Energ. Sci., 8, 535–556, https://doi.org/10.5194/wes-8-535-2023, https://doi.org/10.5194/wes-8-535-2023, 2023
Short summary
Short summary
The study compares small vs. large wind farms regarding the flow and power output with a turbulence-resolving simulation model. It shows that a large wind farm (90 km length) significantly affects the wind direction and that the wind speed is higher in the large wind farm wake. Both wind farms excite atmospheric gravity waves that also affect the power output of the wind farms.
Regis Thedin, Eliot Quon, Matthew Churchfield, and Paul Veers
Wind Energ. Sci., 8, 487–502, https://doi.org/10.5194/wes-8-487-2023, https://doi.org/10.5194/wes-8-487-2023, 2023
Short summary
Short summary
We investigate coherence and correlation and highlight their importance for disciplines like wind energy structural dynamic analysis, in which blade loading and fatigue depend on turbulence structure. We compare coherence estimates to those computed using a model suggested by international standards. We show the differences and highlight additional information that can be gained using large-eddy simulation, further improving analytical coherence models used in synthetic turbulence generators.
Sheng-Lun Tai, Larry K. Berg, Raghavendra Krishnamurthy, Rob Newsom, and Anthony Kirincich
Wind Energ. Sci., 8, 433–448, https://doi.org/10.5194/wes-8-433-2023, https://doi.org/10.5194/wes-8-433-2023, 2023
Short summary
Short summary
Turbulence intensity is critical for wind turbine design and operation as it affects wind power generation efficiency. Turbulence measurements in the marine environment are limited. We use a model to derive turbulence intensity and test how sea surface temperature data may impact the simulated turbulence intensity and atmospheric stability. The model slightly underestimates turbulence, and improved sea surface temperature data reduce the bias. Error with unrealistic mesoscale flow is identified.
Brandon Arthur Lobo, Özge Sinem Özçakmak, Helge Aagaard Madsen, Alois Peter Schaffarczyk, Michael Breuer, and Niels N. Sørensen
Wind Energ. Sci., 8, 303–326, https://doi.org/10.5194/wes-8-303-2023, https://doi.org/10.5194/wes-8-303-2023, 2023
Short summary
Short summary
Results from the DAN-AERO and aerodynamic glove projects provide significant findings. The effects of inflow turbulence on transition and wind turbine blades are compared to computational fluid dynamic simulations. It is found that the transition scenario changes even over a single revolution. The importance of a suitable choice of amplification factor is evident from the simulations. An agreement between the power spectral density plots from the experiment and large-eddy simulations is seen.
Frédéric Blondel
Wind Energ. Sci., 8, 141–147, https://doi.org/10.5194/wes-8-141-2023, https://doi.org/10.5194/wes-8-141-2023, 2023
Short summary
Short summary
Accurate wind farm flow predictions based on analytical wake models are crucial for wind farm design and layout optimization. Wake superposition methods play a key role and remain a substantial source of uncertainty. In the present work, a momentum-conserving superposition method is extended to the superposition of super-Gaussian-type velocity deficit models, allowing the full wake velocity deficit estimation and design of closely packed wind farms.
Abdul Haseeb Syed, Jakob Mann, Andreas Platis, and Jens Bange
Wind Energ. Sci., 8, 125–139, https://doi.org/10.5194/wes-8-125-2023, https://doi.org/10.5194/wes-8-125-2023, 2023
Short summary
Short summary
Wind turbines extract energy from the incoming wind flow, which needs to be recovered. In very large offshore wind farms, the energy is recovered mostly from above the wind farm in a process called entrainment. In this study, we analyzed the effect of atmospheric stability on the entrainment process in large offshore wind farms using measurements recorded by a research aircraft. This is the first time that in situ measurements are used to study the energy recovery process above wind farms.
Cited articles
Almeida, L. B.: Multilayer perceptrons, in: Handbook of Neural Computation, CRC Press, ISBN 9780429142772, 1997. a
Al-Shaikhi, A., Nuha, H., Mohandes, M., Rehman, S., and Adrian, M.: Vertical wind speed extrapolation model using long short-term memory and particle swarm optimization, Energ. Sci. Eng., 10, 4580–4594, https://doi.org/10.1002/ese3.1291, 2022. a, b
Bali, V., Kumar, A., and Gangwar, S.: Deep Learning based Wind Speed Forecasting-A Review, in: IEEE 2019 9th International Conference on Cloud Computing, Data Science & Engineering, 10–11 January 2019, Noida, India, https://doi.org/10.1109/confluence.2019.8776923, 2019. a, b, c
Baquero, L., Torio, H., and Leask, P.: Machine Learning Algorithms for Vertical Wind Speed Data Extrapolation: Comparison and Performance Using Mesoscale and Measured Site Data, Energies, 15, 5518, https://doi.org/10.3390/en15155518, 2022. a, b, c
Beu, C. M. L.: cassiabeu/doi.org-10.5194-wes-2023-104: v1.1, Zenodo [code], https://doi.org/10.5281/zenodo.12168778, 2024. a
Beu, C. M. L. and Landulfo, E.: Turbulence Kinetic Energy Dissipation Rate Estimate for a Low-Level Jet with Doppler Lidar Data: A Case Study, Earth Interact., 26, 112–121, https://doi.org/10.1175/ei-d-20-0027.1, 2022. a, b, c
Breiman, L.: Random Forests, Mach. Learn., 45, 5–32 https://doi.org/10.1023/a:1010933404324, 2001. a
Cheng, C.-H. and Tsai, M.-C.: An Intelligent Time Series Model Based on Hybrid Methodology for Forecasting Concentrations of Significant Air Pollutants, Atmosphere, 13, 1055, https://doi.org/10.3390/atmos13071055, 2022. a
Dalton, A. and Bekker, B.: Exogenous atmospheric variables as wind speed predictors in machine learning, Appl. Energy, 319, 119257, https://doi.org/10.1016/j.apenergy.2022.119257, 2022. a
Efron, B. and Tibshirani, R.: An Introduction to the Bootstrap, Chapman and Hall/CRC, https://doi.org/10.1201/9780429246593, 1994. a
He, J., Yang, H., Zhou, S., Chen, J., and Chen, M.: A Dual-Attention-Mechanism Multi-Channel Convolutional LSTM for Short-Term Wind Speed Prediction, Atmosphere, 14, 71, https://doi.org/10.3390/atmos14010071, 2022. a
Jesemann, A.-S., Matthias, V., Böhner, J., and Bechtel, B.: Using Neural Network NO2-Predictions to Understand Air Quality Changes in Urban Areas – A Case Study in Hamburg, Atmosphere, 13, 1929, https://doi.org/10.3390/atmos13111929, 2022. a
Jiang, H., Wang, X., and Sun, C.: Predicting PM2.5 in the Northeast China Heavy Industrial Zone: A Semi-Supervised Learning with Spatiotemporal Features, Atmosphere, 13, 1744, https://doi.org/10.3390/atmos13111744, 2022. a
Keras: Kerasguide, https://keras.io/api/layers/recurrent_layers/lstm/ (last access: 16 July 2023), 2023. a
Klockow, D. and Targa, H. J.: Performance and results of a six-year German/Brazilian research project in the industrial area of Cubatão/SP Brazil, Pure Appl. Chem., 70, 2287–2293, https://doi.org/10.1351/pac199870122287, 1998. a, b
Liu, B., Ma, X., Guo, J., Li, H., Jin, S., Ma, Y., and Gong, W.: Estimating hub-height wind speed based on a machine learning algorithm: implications for wind energy assessment, Atmos. Chem. Phys., 23, 3181–3193, https://doi.org/10.5194/acp-23-3181-2023, 2023. a
Liu, Y., Cai, J., and Tan, G.: Multi-Level Circulation Pattern Classification Based on the Transfer Learning CNN Network, Atmosphere, 13, 1861, https://doi.org/10.3390/atmos13111861, 2022. a
Medsker, L. and Jain, L. C. (Eds.): Recurrent Neural Networks, CRC Press, https://doi.org/10.1201/9781420049176, 1999. a
Mohandes, M. A. and Rehman, S.: Wind Speed Extrapolation Using Machine Learning Methods and LiDAR Measurements, IEEE Access, 6, 77634–77642, https://doi.org/10.1109/access.2018.2883677, 2018. a, b
Morellato, L. P. C. and Haddad, C. F. B.: Introduction: The Brazilian Atlantic Forest1, Biotropica, 32, 786–792, https://doi.org/10.1111/j.1744-7429.2000.tb00618.x, 2000. a
Mustakim, R., Mamat, M., and Yew, H. T.: Towards On-Site Implementation of Multi-Step Air Pollutant Index Prediction in Malaysia Industrial Area: Comparing the NARX Neural Network and Support Vector Regression, Atmosphere, 13, 1787, https://doi.org/10.3390/atmos13111787, 2022. a
Musyimi, P. K., Sahbeni, G., Timár, G., Weidinger, T., and Székely, B.: Actual Evapotranspiration Estimation Using Sentinel-1 SAR and Sentinel-3 SLSTR Data Combined with a Gradient Boosting Machine Model in Busia County, Western Kenya, Atmosphere, 13, 1927, https://doi.org/10.3390/atmos13111927, 2022. a
Nuha, H., Mohandes, M., Rehman, S., and A-Shaikhi, A.: Vertical wind speed extrapolation using regularized extreme learning machine, FME Trans., 50, 412–421, https://doi.org/10.5937/fme2203412n, 2022. a
O'Malley, T., Bursztein, E., Long, J., et al.: KerasTuner, GitHub [code], https://github.com/keras-team/keras-tuner (last access: 21 July 2023), 2019. a
Pintor, A., Pinto, C., Mendonca, J., Pilao, R., and Pinto, P.: Insights on the use of wind speed vertical extrapolation methods, in: 20th International Conference on Renewable Energies and Power Quality, RE & PQJ, Vigo, Spain, 27–29 July 2022, https://doi.org/10.24084/repqj20.410, 2022. a, b
Ribeiro, F. N., de Oliveira, A. P., Soares, J., de Miranda, R. M., Barlage, M., and Chen, F.: Effect of sea breeze propagation on the urban boundary layer of the metropolitan region of Sao Paulo, Brazil, Atmos. Res., 214, 174–188, https://doi.org/10.1016/j.atmosres.2018.07.015, 2018. a, b
Sánchez, M. P., de Oliveira, A. P., Varona, R. P., Tito, J. V., Codato, G., Ynoue, R. Y., Ribeiro, F. N. D., Filho, E. P. M., and da Silveira, L. C.: Observational Investigation of the Low-Level Jets in the Metropolitan Region of São Paulo, Brazil, Earth Space Sci., 9, e2021EA002190, https://doi.org/10.1029/2021ea002190, 2022. a, b
Schwegmann, S., Faulhaber, J., Pfaffel, S., Yu, Z., Dörenkämper, M., Kersting, K., and Gottschall, J.: Enabling Virtual Met Masts for wind energy applications through machine learning-methods, Energy AI, 11, 100209, https://doi.org/10.1016/j.egyai.2022.100209, 2023. a, b, c
Sherstinsky, A.: Fundamentals of Recurrent Neural Network (RNN) and Long Short-Term Memory (LSTM) network, Physica D, 404, 132306, https://doi.org/10.1016/j.physd.2019.132306, 2020. a
Smagulova, K. and James, A. P.: A survey on LSTM memristive neural network architectures and applications, Eur. Phys. J. Spec. Top., 228, 2313–2324, https://doi.org/10.1140/epjst/e2019-900046-x, 2019. a
Smola, A. J. and Schölkopf, B.: A tutorial on support vector regression, Stat. Comput., https://doi.org/10.1023/b:stco.0000035301.49549.88, 2004. a
Song, Y. and Wang, Y.: Global Wildfire Outlook Forecast with Neural Networks, Remote Sens., 12, 2246, https://doi.org/10.3390/rs12142246, 2020. a
Soria-Ruiz, J., Fernandez-Ordoñez, Y. M., Ambrosio-Ambrosio, J. P., Escalona-Maurice, M. J., Medina-García, G., Sotelo-Ruiz, E. D., and Ramirez-Guzman, M. E.: Flooded Extent and Depth Analysis Using Optical and SAR Remote Sensing with Machine Learning Algorithms, Atmosphere, 13, 1852, https://doi.org/10.3390/atmos13111852, 2022. a
Standen, J., Wilson, C., Vosper, S., and Clark, P.: Prediction of local wind climatology from Met Office models: Virtual Met Mast techniques, Wind Energy, 20, 411–430, https://doi.org/10.1002/we.2013, 2016. a, b
Stull, R. B. (Ed.): An Introduction to Boundary Layer Meteorology, Springer Netherlands, https://doi.org/10.1007/978-94-009-3027-8, 1988. a
Torres, M. E., Colominas, M. A., and Schlotthauer: A complete ensemble empirical mode decomposition with adaptive noise, in: 2011 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 22–27 May 2011, Prague, Czech Republic, 4144–4147, https://doi.org/10.1109/ICASSP.2011.5947265, 2011. a, b
Tukur, A., Chidiebere, O., Shittu, F., and Lawal Abdulrahman, M.: Neural Network Ensemble for Medium Term Forecast of Wind Power Generation: A Review Keyword: Artificial Neural Network, Ensemble technique, Recurrent Neural Network, Deep Learning and Deep Recurrent neural Network, Int. J. Adv. Res. Innov. Idea. Educ., 8, 1856–1865, 2022. a
Türkan, Y. S., Aydoğmuş, H. Y., and Erdal, H.: The prediction of the wind speed at different heights by machine learning methods, Int. J. Optimiz. Control, 6, 179–187, https://doi.org/10.11121/ijocta.01.2016.00315, 2016. a, b
Vassallo, D., Krishnamurthy, R., and Fernando, H. J. S.: Decreasing wind speed extrapolation error via domain-specific feature extraction and selection, Wind Energ. Sci., 5, 959–975, https://doi.org/10.5194/wes-5-959-2020, 2020. a, b, c
Vieira, B. C. and Gramani, M. F.: Serra do Mar: The Most “Tormented” Relief in Brazil, in: World Geomorphological Landscapes, Springer Netherlands, 285–297, https://doi.org/10.1007/978-94-017-8023-0_26, 2015. a
Vieira-Filho, M. S., Lehmann, C., and Fornaro, A.: Influence of local sources and topography on air quality and rainwater composition in Cubatão and São Paulo, Brazil, Atmos. Environ., 101, 200–208, https://doi.org/10.1016/j.atmosenv.2014.11.025, 2015. a
Virtanen, P., Gommers, R., Oliphant, Reddy, T., Cournapeau, Peterson, P., Weckesser, van der Walt, Wilson, J., Millman, Nelson, A. R. J., Jones, Larson, E., Carey, Feng, Y., Moore, Laxalde, D., Perktold, Henriksen, I., Quintero, Archibald, Pedregosa, and SciPy 1.0 Contributors: SciPy 1.0: Fundamental Algorithms for Scientific Computing in Python, Nat. Meth., 17, 261–272, https://doi.org/10.1038/s41592-019-0686-2, 2020. a
Wang, J., Li, Q., and Zeng, B.: Multi-layer cooperative combined forecasting system for short-term wind speed forecasting, Sustain. Energ. Technol. Assess., 43, 100946, https://doi.org/10.1016/j.seta.2020.100946, 2021. a, b
Yu, Y., Si, X., Hu, C., and Zhang, J.: A Review of Recurrent Neural Networks: LSTM Cells and Network Architectures, Neural Comput., 31, 1235–1270, https://doi.org/10.1162/neco_a_01199, 2019. a, b
Zhang, Y., Wang, Y., Zhu, Y., Yang, L., Ge, L., and Luo, C.: Visibility Prediction Based on Machine Learning Algorithms, Atmosphere, 13, 1125, https://doi.org/10.3390/atmos13071125, 2022. a
Zhou, F., Huang, Z., and Zhang, C.: Carbon price forecasting based on CEEMDAN and LSTM, Appl. Energy, 311, 118601, https://doi.org/10.1016/j.apenergy.2022.118601, 2022. a, b
Zhou, J., Feng, J., Zhou, X., Li, Y., and Zhu, F.: Estimating Site-Specific Wind Speeds Using Gridded Data: A Comparison of Multiple Machine Learning Models, Atmosphere, 14, 142, https://doi.org/10.3390/atmos14010142, 2023. a
Short summary
Extrapolating the wind profile for complex terrain through the long short-term memory model outperformed the traditional power law methodology, which due to its universal nature cannot capture local features as the machine-learning methodology does. Moreover, considering the importance of investigating the wind potential and the need for alternative energy sources, it is motivating to find that a short observational campaign can produce better results than the traditional techniques.
Extrapolating the wind profile for complex terrain through the long short-term memory model...
Altmetrics
Final-revised paper
Preprint