View in EDS

Using recurrent neural networks for localized weather prediction with combined use of public airport data and on-site measurements.

Saved in:
Bibliographic Details
Title: Using recurrent neural networks for localized weather prediction with combined use of public airport data and on-site measurements.
Authors: Han, Jung Min1,2 (AUTHOR) jhan2@gsd.harvard.edu, Ang, Yu Qian3 (AUTHOR), Malkawi, Ali1,2 (AUTHOR), Samuelson, Holly W.1,2 (AUTHOR)
Source: Building & Environment. Apr2021, Vol. 192, pN.PAG-N.PAG. 1p.
Subject Terms: *Weather forecasting, *Energy consumption of buildings, *Humidity, *Built environment, Recurrent neural networks, Meteorological stations
Abstract: Weather data is a crucial input for myriad applications in the built environment, including building energy modeling and daylight analysis. Building science practitioners and researchers have been able to select from a variety of weather files, such as Weather Year for Energy Calculation 2 (WYEC2) and the Typical Meteorological Year (TMY). However, commonly used weather files are typically synthesized to represent trends over a relatively longer periods of time, and are often unable to accurately depict climatic conditions that result from local contexts, such as the heat island effect, wind flow, even local temperature and relative humidity. This results in discrepancies in building performance simulations. This study proposes a methodology using recurrent neural networks to generate synthetic localized weather data that are significantly more accurate and representative of local conditions than standard weather files. The predictions were validated against actual on-site measurements, and achieved a low mean square error of 2.96 and over 185% improvement in validation accuracy. Overall, the performance of selected models has shown over 100% improvements in test accuracy compared with standard weather files and weather station data at the nearest airport. The proposed methodology can be used to morph generic weather files to accurately represent localized conditions, or generate localized data for a longer time span with only a subset of data available/collected. This is useful for downstream built environment applications, especially building energy modeling, since representative weather data capturing trends of temperature and other variables will result in enhanced accuracies of the building energy models. The method can also be used in urban analysis pipelines to enhance resilience against climate change. • Recurrent neural networks (RNN) can model localized weather accurately. • Different recurrent neutral network architectures are analyzed. • Gated Recurrent Units (GRU) performed best in the study with low mean squared error. • Our model has more than 185% improvement over generic weather data from airports. • The study discusses impact of weather files on building performance simulations. [ABSTRACT FROM AUTHOR]
Copyright of Building & Environment is the property of Pergamon Press - An Imprint of Elsevier Science and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database: GreenFILE
FullText Text:
  Availability: 0
Header DbId: 8gh
DbLabel: GreenFILE
An: 148776195
AccessLevel: 6
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: Using recurrent neural networks for localized weather prediction with combined use of public airport data and on-site measurements.
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Han%2C+Jung+Min%22">Han, Jung Min</searchLink><relatesTo>1,2</relatesTo> (AUTHOR)<i> jhan2@gsd.harvard.edu</i><br /><searchLink fieldCode="AR" term="%22Ang%2C+Yu+Qian%22">Ang, Yu Qian</searchLink><relatesTo>3</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Malkawi%2C+Ali%22">Malkawi, Ali</searchLink><relatesTo>1,2</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Samuelson%2C+Holly+W%2E%22">Samuelson, Holly W.</searchLink><relatesTo>1,2</relatesTo> (AUTHOR)
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="JN" term="%22Building+%26+Environment%22">Building & Environment</searchLink>. Apr2021, Vol. 192, pN.PAG-N.PAG. 1p.
– Name: Subject
  Label: Subject Terms
  Group: Su
  Data: *<searchLink fieldCode="DE" term="%22Weather+forecasting%22">Weather forecasting</searchLink><br />*<searchLink fieldCode="DE" term="%22Energy+consumption+of+buildings%22">Energy consumption of buildings</searchLink><br />*<searchLink fieldCode="DE" term="%22Humidity%22">Humidity</searchLink><br />*<searchLink fieldCode="DE" term="%22Built+environment%22">Built environment</searchLink><br /><searchLink fieldCode="DE" term="%22Recurrent+neural+networks%22">Recurrent neural networks</searchLink><br /><searchLink fieldCode="DE" term="%22Meteorological+stations%22">Meteorological stations</searchLink>
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: Weather data is a crucial input for myriad applications in the built environment, including building energy modeling and daylight analysis. Building science practitioners and researchers have been able to select from a variety of weather files, such as Weather Year for Energy Calculation 2 (WYEC2) and the Typical Meteorological Year (TMY). However, commonly used weather files are typically synthesized to represent trends over a relatively longer periods of time, and are often unable to accurately depict climatic conditions that result from local contexts, such as the heat island effect, wind flow, even local temperature and relative humidity. This results in discrepancies in building performance simulations. This study proposes a methodology using recurrent neural networks to generate synthetic localized weather data that are significantly more accurate and representative of local conditions than standard weather files. The predictions were validated against actual on-site measurements, and achieved a low mean square error of 2.96 and over 185% improvement in validation accuracy. Overall, the performance of selected models has shown over 100% improvements in test accuracy compared with standard weather files and weather station data at the nearest airport. The proposed methodology can be used to morph generic weather files to accurately represent localized conditions, or generate localized data for a longer time span with only a subset of data available/collected. This is useful for downstream built environment applications, especially building energy modeling, since representative weather data capturing trends of temperature and other variables will result in enhanced accuracies of the building energy models. The method can also be used in urban analysis pipelines to enhance resilience against climate change. • Recurrent neural networks (RNN) can model localized weather accurately. • Different recurrent neutral network architectures are analyzed. • Gated Recurrent Units (GRU) performed best in the study with low mean squared error. • Our model has more than 185% improvement over generic weather data from airports. • The study discusses impact of weather files on building performance simulations. [ABSTRACT FROM AUTHOR]
– Name: AbstractSuppliedCopyright
  Label:
  Group: Ab
  Data: <i>Copyright of Building & Environment is the property of Pergamon Press - An Imprint of Elsevier Science and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract.</i> (Copyright applies to all Abstracts.)
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=8gh&AN=148776195
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1016/j.buildenv.2021.107601
    Languages:
      – Code: eng
        Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 1
        StartPage: N.PAG
    Subjects:
      – SubjectFull: Weather forecasting
        Type: general
      – SubjectFull: Energy consumption of buildings
        Type: general
      – SubjectFull: Humidity
        Type: general
      – SubjectFull: Built environment
        Type: general
      – SubjectFull: Recurrent neural networks
        Type: general
      – SubjectFull: Meteorological stations
        Type: general
    Titles:
      – TitleFull: Using recurrent neural networks for localized weather prediction with combined use of public airport data and on-site measurements.
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Han, Jung Min
      – PersonEntity:
          Name:
            NameFull: Ang, Yu Qian
      – PersonEntity:
          Name:
            NameFull: Malkawi, Ali
      – PersonEntity:
          Name:
            NameFull: Samuelson, Holly W.
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 04
              Text: Apr2021
              Type: published
              Y: 2021
          Identifiers:
            – Type: issn-print
              Value: 03601323
          Numbering:
            – Type: volume
              Value: 192
          Titles:
            – TitleFull: Building & Environment
              Type: main
ResultId 1