Learning the structure of wind: A data-driven nonlocal turbulence model for the atmospheric boundary layer
We develop a novel data-driven approach to modeling the atmospheric boundary layer. This approach leads to a nonlocal, anisotropic synthetic turbulence model which we refer to as the deep rapid distortion (DRD) model. Our approach relies on an operator regression problem that characterizes the best...
Gespeichert in:
Veröffentlicht in: | Physics of fluids (1994) 2021-09, Vol.33 (9) |
---|---|
Hauptverfasser: | , , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
container_end_page | |
---|---|
container_issue | 9 |
container_start_page | |
container_title | Physics of fluids (1994) |
container_volume | 33 |
creator | Keith, B. Khristenko, U. Wohlmuth, B. |
description | We develop a novel data-driven approach to modeling the atmospheric boundary layer. This approach leads to a nonlocal, anisotropic synthetic turbulence model which we refer to as the deep rapid distortion (DRD) model. Our approach relies on an operator regression problem that characterizes the best fitting candidate in a general family of nonlocal covariance kernels parameterized in part by a neural network. This family of covariance kernels is expressed in Fourier space and is obtained from approximate solutions to the Navier–Stokes equations at very high Reynolds numbers. Each member of the family incorporates important physical properties such as mass conservation and a realistic energy cascade. The DRD model can be calibrated with noisy data from field experiments. After calibration, the model can be used to generate synthetic turbulent velocity fields. To this end, we provide a new numerical method based on domain decomposition which delivers scalable, memory-efficient turbulence generation with the DRD model as well as others. We demonstrate the robustness of our approach with both filtered and noisy data coming from the 1968 Air Force Cambridge Research Laboratory Kansas experiments. Using these data, we witness exceptional accuracy with the DRD model, especially when compared to the International Electrotechnical Commission standard. |
doi_str_mv | 10.1063/5.0064394 |
format | Article |
fullrecord | <record><control><sourceid>proquest_osti_</sourceid><recordid>TN_cdi_osti_scitechconnect_1834490</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2570606024</sourcerecordid><originalsourceid>FETCH-LOGICAL-c319t-5bffd431b2cd1ecb32e7dd79a9e75974c4a4b38d87c19a1e3c84ea7e9bd9d75f3</originalsourceid><addsrcrecordid>eNp90DtPwzAQAGALgUQpDPwDCyaQUuzYsWM2VPGSKrHAbDn2haZK7WI7oP57UtoZ3XA3fLoXQpeUzCgR7K6aESI4U_wITSipVSGFEMe7WpJCCEZP0VlKK0IIU6WYoNUCTPSd_8R5CTjlONg8RMChxT-dd_f4ATuTTeFi9w0e--D7YE2PR9QMPXgLeB0c9LgN8a-FyeuQNkuIncVNGLwzcYt7s4V4jk5a0ye4OOQp-nh6fJ-_FIu359f5w6KwjKpcVE3bOs5oU1pHwTasBOmcVEaBrJTklhvesNrV0lJlKDBbczASVOOUk1XLpuhq3zek3Olkuwx2aYP3YLOmNeNckRFd79Emhq8BUtarMEQ_7qXLShIxRslHdbNXNoaUIrR6E7v1eJGmRO_-rSt9-Pdob_d2N9HkLvh_8C-tQYDU</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2570606024</pqid></control><display><type>article</type><title>Learning the structure of wind: A data-driven nonlocal turbulence model for the atmospheric boundary layer</title><source>AIP Journals Complete</source><source>Alma/SFX Local Collection</source><creator>Keith, B. ; Khristenko, U. ; Wohlmuth, B.</creator><creatorcontrib>Keith, B. ; Khristenko, U. ; Wohlmuth, B. ; Lawrence Livermore National Laboratory (LLNL), Livermore, CA (United States)</creatorcontrib><description>We develop a novel data-driven approach to modeling the atmospheric boundary layer. This approach leads to a nonlocal, anisotropic synthetic turbulence model which we refer to as the deep rapid distortion (DRD) model. Our approach relies on an operator regression problem that characterizes the best fitting candidate in a general family of nonlocal covariance kernels parameterized in part by a neural network. This family of covariance kernels is expressed in Fourier space and is obtained from approximate solutions to the Navier–Stokes equations at very high Reynolds numbers. Each member of the family incorporates important physical properties such as mass conservation and a realistic energy cascade. The DRD model can be calibrated with noisy data from field experiments. After calibration, the model can be used to generate synthetic turbulent velocity fields. To this end, we provide a new numerical method based on domain decomposition which delivers scalable, memory-efficient turbulence generation with the DRD model as well as others. We demonstrate the robustness of our approach with both filtered and noisy data coming from the 1968 Air Force Cambridge Research Laboratory Kansas experiments. Using these data, we witness exceptional accuracy with the DRD model, especially when compared to the International Electrotechnical Commission standard.</description><identifier>ISSN: 1070-6631</identifier><identifier>EISSN: 1089-7666</identifier><identifier>DOI: 10.1063/5.0064394</identifier><identifier>CODEN: PHFLE6</identifier><language>eng</language><publisher>Melville: American Institute of Physics</publisher><subject>Artificial neural networks ; Atmospheric boundary layer ; Atmospheric models ; Computational fluid dynamics ; Covariance ; Domain decomposition methods ; Fluid dynamics ; Fluid flow ; High Reynolds number ; Kernels ; Navier Stokes equations ; Neural networks ; Numerical methods ; Physical properties ; Physics ; Robustness (mathematics) ; Turbulence models ; Turbulence simulations ; Turbulence theory and modelling ; Turbulent flows ; Velocity distribution ; WIND ENERGY</subject><ispartof>Physics of fluids (1994), 2021-09, Vol.33 (9)</ispartof><rights>Author(s)</rights><rights>2021 Author(s). Published under an exclusive license by AIP Publishing.</rights><lds50>peer_reviewed</lds50><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c319t-5bffd431b2cd1ecb32e7dd79a9e75974c4a4b38d87c19a1e3c84ea7e9bd9d75f3</citedby><cites>FETCH-LOGICAL-c319t-5bffd431b2cd1ecb32e7dd79a9e75974c4a4b38d87c19a1e3c84ea7e9bd9d75f3</cites><orcidid>0000-0002-4573-0336 ; 0000-0001-6908-6015 ; 0000-0002-6969-6857 ; 0000000245730336 ; 0000000169086015 ; 0000000269696857</orcidid></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>230,314,778,782,792,883,4500,27911,27912</link.rule.ids><backlink>$$Uhttps://www.osti.gov/servlets/purl/1834490$$D View this record in Osti.gov$$Hfree_for_read</backlink></links><search><creatorcontrib>Keith, B.</creatorcontrib><creatorcontrib>Khristenko, U.</creatorcontrib><creatorcontrib>Wohlmuth, B.</creatorcontrib><creatorcontrib>Lawrence Livermore National Laboratory (LLNL), Livermore, CA (United States)</creatorcontrib><title>Learning the structure of wind: A data-driven nonlocal turbulence model for the atmospheric boundary layer</title><title>Physics of fluids (1994)</title><description>We develop a novel data-driven approach to modeling the atmospheric boundary layer. This approach leads to a nonlocal, anisotropic synthetic turbulence model which we refer to as the deep rapid distortion (DRD) model. Our approach relies on an operator regression problem that characterizes the best fitting candidate in a general family of nonlocal covariance kernels parameterized in part by a neural network. This family of covariance kernels is expressed in Fourier space and is obtained from approximate solutions to the Navier–Stokes equations at very high Reynolds numbers. Each member of the family incorporates important physical properties such as mass conservation and a realistic energy cascade. The DRD model can be calibrated with noisy data from field experiments. After calibration, the model can be used to generate synthetic turbulent velocity fields. To this end, we provide a new numerical method based on domain decomposition which delivers scalable, memory-efficient turbulence generation with the DRD model as well as others. We demonstrate the robustness of our approach with both filtered and noisy data coming from the 1968 Air Force Cambridge Research Laboratory Kansas experiments. Using these data, we witness exceptional accuracy with the DRD model, especially when compared to the International Electrotechnical Commission standard.</description><subject>Artificial neural networks</subject><subject>Atmospheric boundary layer</subject><subject>Atmospheric models</subject><subject>Computational fluid dynamics</subject><subject>Covariance</subject><subject>Domain decomposition methods</subject><subject>Fluid dynamics</subject><subject>Fluid flow</subject><subject>High Reynolds number</subject><subject>Kernels</subject><subject>Navier Stokes equations</subject><subject>Neural networks</subject><subject>Numerical methods</subject><subject>Physical properties</subject><subject>Physics</subject><subject>Robustness (mathematics)</subject><subject>Turbulence models</subject><subject>Turbulence simulations</subject><subject>Turbulence theory and modelling</subject><subject>Turbulent flows</subject><subject>Velocity distribution</subject><subject>WIND ENERGY</subject><issn>1070-6631</issn><issn>1089-7666</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2021</creationdate><recordtype>article</recordtype><recordid>eNp90DtPwzAQAGALgUQpDPwDCyaQUuzYsWM2VPGSKrHAbDn2haZK7WI7oP57UtoZ3XA3fLoXQpeUzCgR7K6aESI4U_wITSipVSGFEMe7WpJCCEZP0VlKK0IIU6WYoNUCTPSd_8R5CTjlONg8RMChxT-dd_f4ATuTTeFi9w0e--D7YE2PR9QMPXgLeB0c9LgN8a-FyeuQNkuIncVNGLwzcYt7s4V4jk5a0ye4OOQp-nh6fJ-_FIu359f5w6KwjKpcVE3bOs5oU1pHwTasBOmcVEaBrJTklhvesNrV0lJlKDBbczASVOOUk1XLpuhq3zek3Olkuwx2aYP3YLOmNeNckRFd79Emhq8BUtarMEQ_7qXLShIxRslHdbNXNoaUIrR6E7v1eJGmRO_-rSt9-Pdob_d2N9HkLvh_8C-tQYDU</recordid><startdate>20210901</startdate><enddate>20210901</enddate><creator>Keith, B.</creator><creator>Khristenko, U.</creator><creator>Wohlmuth, B.</creator><general>American Institute of Physics</general><general>American Institute of Physics (AIP)</general><scope>AAYXX</scope><scope>CITATION</scope><scope>8FD</scope><scope>H8D</scope><scope>L7M</scope><scope>OIOZB</scope><scope>OTOTI</scope><orcidid>https://orcid.org/0000-0002-4573-0336</orcidid><orcidid>https://orcid.org/0000-0001-6908-6015</orcidid><orcidid>https://orcid.org/0000-0002-6969-6857</orcidid><orcidid>https://orcid.org/0000000245730336</orcidid><orcidid>https://orcid.org/0000000169086015</orcidid><orcidid>https://orcid.org/0000000269696857</orcidid></search><sort><creationdate>20210901</creationdate><title>Learning the structure of wind: A data-driven nonlocal turbulence model for the atmospheric boundary layer</title><author>Keith, B. ; Khristenko, U. ; Wohlmuth, B.</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c319t-5bffd431b2cd1ecb32e7dd79a9e75974c4a4b38d87c19a1e3c84ea7e9bd9d75f3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2021</creationdate><topic>Artificial neural networks</topic><topic>Atmospheric boundary layer</topic><topic>Atmospheric models</topic><topic>Computational fluid dynamics</topic><topic>Covariance</topic><topic>Domain decomposition methods</topic><topic>Fluid dynamics</topic><topic>Fluid flow</topic><topic>High Reynolds number</topic><topic>Kernels</topic><topic>Navier Stokes equations</topic><topic>Neural networks</topic><topic>Numerical methods</topic><topic>Physical properties</topic><topic>Physics</topic><topic>Robustness (mathematics)</topic><topic>Turbulence models</topic><topic>Turbulence simulations</topic><topic>Turbulence theory and modelling</topic><topic>Turbulent flows</topic><topic>Velocity distribution</topic><topic>WIND ENERGY</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Keith, B.</creatorcontrib><creatorcontrib>Khristenko, U.</creatorcontrib><creatorcontrib>Wohlmuth, B.</creatorcontrib><creatorcontrib>Lawrence Livermore National Laboratory (LLNL), Livermore, CA (United States)</creatorcontrib><collection>CrossRef</collection><collection>Technology Research Database</collection><collection>Aerospace Database</collection><collection>Advanced Technologies Database with Aerospace</collection><collection>OSTI.GOV - Hybrid</collection><collection>OSTI.GOV</collection><jtitle>Physics of fluids (1994)</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Keith, B.</au><au>Khristenko, U.</au><au>Wohlmuth, B.</au><aucorp>Lawrence Livermore National Laboratory (LLNL), Livermore, CA (United States)</aucorp><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Learning the structure of wind: A data-driven nonlocal turbulence model for the atmospheric boundary layer</atitle><jtitle>Physics of fluids (1994)</jtitle><date>2021-09-01</date><risdate>2021</risdate><volume>33</volume><issue>9</issue><issn>1070-6631</issn><eissn>1089-7666</eissn><coden>PHFLE6</coden><abstract>We develop a novel data-driven approach to modeling the atmospheric boundary layer. This approach leads to a nonlocal, anisotropic synthetic turbulence model which we refer to as the deep rapid distortion (DRD) model. Our approach relies on an operator regression problem that characterizes the best fitting candidate in a general family of nonlocal covariance kernels parameterized in part by a neural network. This family of covariance kernels is expressed in Fourier space and is obtained from approximate solutions to the Navier–Stokes equations at very high Reynolds numbers. Each member of the family incorporates important physical properties such as mass conservation and a realistic energy cascade. The DRD model can be calibrated with noisy data from field experiments. After calibration, the model can be used to generate synthetic turbulent velocity fields. To this end, we provide a new numerical method based on domain decomposition which delivers scalable, memory-efficient turbulence generation with the DRD model as well as others. We demonstrate the robustness of our approach with both filtered and noisy data coming from the 1968 Air Force Cambridge Research Laboratory Kansas experiments. Using these data, we witness exceptional accuracy with the DRD model, especially when compared to the International Electrotechnical Commission standard.</abstract><cop>Melville</cop><pub>American Institute of Physics</pub><doi>10.1063/5.0064394</doi><tpages>10</tpages><orcidid>https://orcid.org/0000-0002-4573-0336</orcidid><orcidid>https://orcid.org/0000-0001-6908-6015</orcidid><orcidid>https://orcid.org/0000-0002-6969-6857</orcidid><orcidid>https://orcid.org/0000000245730336</orcidid><orcidid>https://orcid.org/0000000169086015</orcidid><orcidid>https://orcid.org/0000000269696857</orcidid><oa>free_for_read</oa></addata></record> |
fulltext | fulltext |
identifier | ISSN: 1070-6631 |
ispartof | Physics of fluids (1994), 2021-09, Vol.33 (9) |
issn | 1070-6631 1089-7666 |
language | eng |
recordid | cdi_osti_scitechconnect_1834490 |
source | AIP Journals Complete; Alma/SFX Local Collection |
subjects | Artificial neural networks Atmospheric boundary layer Atmospheric models Computational fluid dynamics Covariance Domain decomposition methods Fluid dynamics Fluid flow High Reynolds number Kernels Navier Stokes equations Neural networks Numerical methods Physical properties Physics Robustness (mathematics) Turbulence models Turbulence simulations Turbulence theory and modelling Turbulent flows Velocity distribution WIND ENERGY |
title | Learning the structure of wind: A data-driven nonlocal turbulence model for the atmospheric boundary layer |
url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-15T12%3A41%3A18IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_osti_&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Learning%20the%20structure%20of%20wind:%20A%20data-driven%20nonlocal%20turbulence%20model%20for%20the%20atmospheric%20boundary%20layer&rft.jtitle=Physics%20of%20fluids%20(1994)&rft.au=Keith,%20B.&rft.aucorp=Lawrence%20Livermore%20National%20Laboratory%20(LLNL),%20Livermore,%20CA%20(United%20States)&rft.date=2021-09-01&rft.volume=33&rft.issue=9&rft.issn=1070-6631&rft.eissn=1089-7666&rft.coden=PHFLE6&rft_id=info:doi/10.1063/5.0064394&rft_dat=%3Cproquest_osti_%3E2570606024%3C/proquest_osti_%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2570606024&rft_id=info:pmid/&rfr_iscdi=true |