Academic Journal
View in EDS
Linked Full Text

Spatiotemporal inhomogeneity of accuracy degradation in AI weather forecast foundation models: A GNSS perspective

Bibliographic Details
Title: Spatiotemporal inhomogeneity of accuracy degradation in AI weather forecast foundation models: A GNSS perspective
Authors: Junsheng Ding, Wu Chen, Junping Chen, Jungang Wang, Yize Zhang, Lei Bai, Yuyan Wang, Xiaolong Mi, Tong Liu, Duojie Weng
Source: International Journal of Applied Earth Observations and Geoinformation, Vol 139, Iss , Pp 104473- (2025)
Publisher Information: Elsevier, 2025.
Publication Year: 2025
Collection: LCC:Physical geography
LCC:Environmental sciences
Subject Terms: Foundation models, GNSS tropospheric delay, Spatiotemporal inhomogeneity, Accuracy degradation, Weather forecast, Physical geography, GB3-5030, Environmental sciences, GE1-350
More Details: The artificial intelligence (AI) weather forecast foundation models can infer and generate precise global atmospheric state forecasts on the user’s device and with speed over 10,000 times faster than the operational Integrated Forecasting System (IFS), and it is making increasingly significant contributions to geodetic applications represented by the Global Navigation Satellite System (GNSS). However, existing studies on the investigation of these AI models are typically carried out by concentrating on specific one or several meteorological events in certain regions or by comparison with physical models, and the evaluation results obtained in this manner are not comprehensive and universal. Additionally, we find that the results obtained by the foundation models through the “rollout” method for forecasting are not uniform in terms of time and space. This temporal and spatial inhomogeneity of accuracy and accuracy degradation are related to AI algorithms and attributes of training data, etc., but these characteristics have not been thoroughly explored and analyzed. In this study, we obtained the global forecast results of foundation models for 2022 and subsequently derived the GNSS tropospheric delay through numerical integration. We calculated the mean deviation, mean absolute error, and root mean square error of these data. Using these metrics, we analyzed the spatiotemporal inhomogeneity in the accuracy degradation of foundation models, represented by Huawei Cloud Pangu-Weather, Google DeepMind GraphCast, and Shanghai AI Lab FengWu. We evaluated how this inhomogeneity changes with forecast time and identified the best-performing models across different regions and forecast durations. From the results, we find that taking topography into account when training the model enhances its accuracy at high altitudes, and the facilitating influence between the high related atmospheric variables such as precipitation and water vapor. The contributions of this study are twofold: it serves as a valuable reference for geodetic and remote sensing users employing foundational models, and offers insights and case supports for AI practitioners aiming to develop more accurate models for weather forecasting.
Document Type: article
File Description: electronic resource
Language: English
ISSN: 1569-8432
Relation: http://www.sciencedirect.com/science/article/pii/S1569843225001207; https://doaj.org/toc/1569-8432
DOI: 10.1016/j.jag.2025.104473
Access URL: https://doaj.org/article/4942496a874c4596882c398f73da34f7
Accession Number: edsdoj.4942496a874c4596882c398f73da34f7
Database: Directory of Open Access Journals
FullText Links:
  – Type: other
    Url: https://resolver.ebsco.com:443/public/rma-ftfapi/ejs/direct?AccessToken=4022AC51B5EA156FEE17&Show=Object
Text:
  Availability: 0
CustomLinks:
  – Url: https://www.doi.org/10.1016/j.jag.2025.104473?
    Name: ScienceDirect (all content)-s8985755
    Category: fullText
    Text: View record from ScienceDirect
    MouseOverText: View record from ScienceDirect
  – Url: https://resolver.ebsco.com/c/xy5jbn/result?sid=EBSCO:edsdoj&genre=article&issn=15698432&ISBN=&volume=139&issue=104473-&date=20250501&spage=&pages=&title=International Journal of Applied Earth Observations and Geoinformation&atitle=Spatiotemporal%20inhomogeneity%20of%20accuracy%20degradation%20in%20AI%20weather%20forecast%20foundation%20models%3A%20A%20GNSS%20perspective&aulast=Junsheng%20Ding&id=DOI:10.1016/j.jag.2025.104473
    Name: Full Text Finder (for New FTF UI) (s8985755)
    Category: fullText
    Text: Find It @ SCU Libraries
    MouseOverText: Find It @ SCU Libraries
  – Url: https://doaj.org/article/4942496a874c4596882c398f73da34f7
    Name: EDS - DOAJ (s8985755)
    Category: fullText
    Text: View record from DOAJ
    MouseOverText: View record from DOAJ
Header DbId: edsdoj
DbLabel: Directory of Open Access Journals
An: edsdoj.4942496a874c4596882c398f73da34f7
RelevancyScore: 1044
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 1044.00451660156
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: Spatiotemporal inhomogeneity of accuracy degradation in AI weather forecast foundation models: A GNSS perspective
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Junsheng+Ding%22">Junsheng Ding</searchLink><br /><searchLink fieldCode="AR" term="%22Wu+Chen%22">Wu Chen</searchLink><br /><searchLink fieldCode="AR" term="%22Junping+Chen%22">Junping Chen</searchLink><br /><searchLink fieldCode="AR" term="%22Jungang+Wang%22">Jungang Wang</searchLink><br /><searchLink fieldCode="AR" term="%22Yize+Zhang%22">Yize Zhang</searchLink><br /><searchLink fieldCode="AR" term="%22Lei+Bai%22">Lei Bai</searchLink><br /><searchLink fieldCode="AR" term="%22Yuyan+Wang%22">Yuyan Wang</searchLink><br /><searchLink fieldCode="AR" term="%22Xiaolong+Mi%22">Xiaolong Mi</searchLink><br /><searchLink fieldCode="AR" term="%22Tong+Liu%22">Tong Liu</searchLink><br /><searchLink fieldCode="AR" term="%22Duojie+Weng%22">Duojie Weng</searchLink>
– Name: TitleSource
  Label: Source
  Group: Src
  Data: International Journal of Applied Earth Observations and Geoinformation, Vol 139, Iss , Pp 104473- (2025)
– Name: Publisher
  Label: Publisher Information
  Group: PubInfo
  Data: Elsevier, 2025.
– Name: DatePubCY
  Label: Publication Year
  Group: Date
  Data: 2025
– Name: Subset
  Label: Collection
  Group: HoldingsInfo
  Data: LCC:Physical geography<br />LCC:Environmental sciences
– Name: Subject
  Label: Subject Terms
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Foundation+models%22">Foundation models</searchLink><br /><searchLink fieldCode="DE" term="%22GNSS+tropospheric+delay%22">GNSS tropospheric delay</searchLink><br /><searchLink fieldCode="DE" term="%22Spatiotemporal+inhomogeneity%22">Spatiotemporal inhomogeneity</searchLink><br /><searchLink fieldCode="DE" term="%22Accuracy+degradation%22">Accuracy degradation</searchLink><br /><searchLink fieldCode="DE" term="%22Weather+forecast%22">Weather forecast</searchLink><br /><searchLink fieldCode="DE" term="%22Physical+geography%22">Physical geography</searchLink><br /><searchLink fieldCode="DE" term="%22GB3-5030%22">GB3-5030</searchLink><br /><searchLink fieldCode="DE" term="%22Environmental+sciences%22">Environmental sciences</searchLink><br /><searchLink fieldCode="DE" term="%22GE1-350%22">GE1-350</searchLink>
– Name: Abstract
  Label: Description
  Group: Ab
  Data: The artificial intelligence (AI) weather forecast foundation models can infer and generate precise global atmospheric state forecasts on the user’s device and with speed over 10,000 times faster than the operational Integrated Forecasting System (IFS), and it is making increasingly significant contributions to geodetic applications represented by the Global Navigation Satellite System (GNSS). However, existing studies on the investigation of these AI models are typically carried out by concentrating on specific one or several meteorological events in certain regions or by comparison with physical models, and the evaluation results obtained in this manner are not comprehensive and universal. Additionally, we find that the results obtained by the foundation models through the “rollout” method for forecasting are not uniform in terms of time and space. This temporal and spatial inhomogeneity of accuracy and accuracy degradation are related to AI algorithms and attributes of training data, etc., but these characteristics have not been thoroughly explored and analyzed. In this study, we obtained the global forecast results of foundation models for 2022 and subsequently derived the GNSS tropospheric delay through numerical integration. We calculated the mean deviation, mean absolute error, and root mean square error of these data. Using these metrics, we analyzed the spatiotemporal inhomogeneity in the accuracy degradation of foundation models, represented by Huawei Cloud Pangu-Weather, Google DeepMind GraphCast, and Shanghai AI Lab FengWu. We evaluated how this inhomogeneity changes with forecast time and identified the best-performing models across different regions and forecast durations. From the results, we find that taking topography into account when training the model enhances its accuracy at high altitudes, and the facilitating influence between the high related atmospheric variables such as precipitation and water vapor. The contributions of this study are twofold: it serves as a valuable reference for geodetic and remote sensing users employing foundational models, and offers insights and case supports for AI practitioners aiming to develop more accurate models for weather forecasting.
– Name: TypeDocument
  Label: Document Type
  Group: TypDoc
  Data: article
– Name: Format
  Label: File Description
  Group: SrcInfo
  Data: electronic resource
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: ISSN
  Label: ISSN
  Group: ISSN
  Data: 1569-8432
– Name: NoteTitleSource
  Label: Relation
  Group: SrcInfo
  Data: http://www.sciencedirect.com/science/article/pii/S1569843225001207; https://doaj.org/toc/1569-8432
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.1016/j.jag.2025.104473
– Name: URL
  Label: Access URL
  Group: URL
  Data: <link linkTarget="URL" linkTerm="https://doaj.org/article/4942496a874c4596882c398f73da34f7" linkWindow="_blank">https://doaj.org/article/4942496a874c4596882c398f73da34f7</link>
– Name: AN
  Label: Accession Number
  Group: ID
  Data: edsdoj.4942496a874c4596882c398f73da34f7
PLink https://login.libproxy.scu.edu/login?url=https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&scope=site&db=edsdoj&AN=edsdoj.4942496a874c4596882c398f73da34f7
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1016/j.jag.2025.104473
    Languages:
      – Text: English
    Subjects:
      – SubjectFull: Foundation models
        Type: general
      – SubjectFull: GNSS tropospheric delay
        Type: general
      – SubjectFull: Spatiotemporal inhomogeneity
        Type: general
      – SubjectFull: Accuracy degradation
        Type: general
      – SubjectFull: Weather forecast
        Type: general
      – SubjectFull: Physical geography
        Type: general
      – SubjectFull: GB3-5030
        Type: general
      – SubjectFull: Environmental sciences
        Type: general
      – SubjectFull: GE1-350
        Type: general
    Titles:
      – TitleFull: Spatiotemporal inhomogeneity of accuracy degradation in AI weather forecast foundation models: A GNSS perspective
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Junsheng Ding
      – PersonEntity:
          Name:
            NameFull: Wu Chen
      – PersonEntity:
          Name:
            NameFull: Junping Chen
      – PersonEntity:
          Name:
            NameFull: Jungang Wang
      – PersonEntity:
          Name:
            NameFull: Yize Zhang
      – PersonEntity:
          Name:
            NameFull: Lei Bai
      – PersonEntity:
          Name:
            NameFull: Yuyan Wang
      – PersonEntity:
          Name:
            NameFull: Xiaolong Mi
      – PersonEntity:
          Name:
            NameFull: Tong Liu
      – PersonEntity:
          Name:
            NameFull: Duojie Weng
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 05
              Type: published
              Y: 2025
          Identifiers:
            – Type: issn-print
              Value: 15698432
          Numbering:
            – Type: volume
              Value: 139
            – Type: issue
              Value: 104473-
          Titles:
            – TitleFull: International Journal of Applied Earth Observations and Geoinformation
              Type: main
ResultId 1