Decoding the Spatiotemporal Complexities of the Permafrost Carbon Feedback With Multimodal Ensemble Learning
Abstract Complex nonlinear relationships exist between the permafrost thermal state, active layer thickness, and terrestrial carbon cycle dynamics. In Arctic and boreal Alaska, significant uncertainties characterize the spatiotemporal rate and magnitude of permafrost degradation and the permafrost c...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2025-03-01
|
| Series: | Journal of Geophysical Research: Machine Learning and Computation |
| Subjects: | |
| Online Access: | https://doi.org/10.1029/2024JH000402 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850206725719719936 |
|---|---|
| author | B. A. Gay N. J. Pastick J. D. Watts A. H. Armstrong K. R. Miner C. E. Miller |
| author_facet | B. A. Gay N. J. Pastick J. D. Watts A. H. Armstrong K. R. Miner C. E. Miller |
| author_sort | B. A. Gay |
| collection | DOAJ |
| description | Abstract Complex nonlinear relationships exist between the permafrost thermal state, active layer thickness, and terrestrial carbon cycle dynamics. In Arctic and boreal Alaska, significant uncertainties characterize the spatiotemporal rate and magnitude of permafrost degradation and the permafrost carbon feedback, with increasing recognition of the importance of thawing mechanisms. The challenges of monitoring sub‐surface phenomena with remote sensing technology further complicate the issue. There is an urgent need to understand how and to what extent thawing permafrost destabilizes the carbon balance in Alaska and to characterize the feedback involved. In this research, we use our artificial intelligence‐driven model GeoCryoAI to quantify permafrost carbon dynamics in Alaska. The GeoCryoAI model uses a hybridized process‐constrained ensemble learning framework to simultaneously ingest, scale, and analyze in situ measurements, remote sensing observations, and process‐based modeling outputs with disparate spatiotemporal sampling and data densities. We evaluated prior naïve (a) persistence and (b) teacher forcing approaches relative to (c) time‐delayed GeoCryoAI simulations, yielding the following error metrics (RMSE) for active layer thickness (ALT), methane (CH4), and carbon dioxide (CO2), respectively: 1.997, 1.327, 1.007 cm [1963–2022]; 0.884, 0.715, 0.694 nmol CH4km−2 month−1 [1994–2022]; 1.906, 0.697, 0.213 µmol CO2km−2 month−1 [1994–2022]. Our approach overcomes traditional model inefficiencies and resolves spatiotemporal disparities. GeoCryoAI captures abrupt and persistent changes while introducing a novel methodology for assimilating contemporaneous information at various scales. We describe GeoCryoAI, the methodology, our results, and plans for future applications. |
| format | Article |
| id | doaj-art-02332f29fe2f45f9891fe8ffaba9c30b |
| institution | OA Journals |
| issn | 2993-5210 |
| language | English |
| publishDate | 2025-03-01 |
| publisher | Wiley |
| record_format | Article |
| series | Journal of Geophysical Research: Machine Learning and Computation |
| spelling | doaj-art-02332f29fe2f45f9891fe8ffaba9c30b2025-08-20T02:10:42ZengWileyJournal of Geophysical Research: Machine Learning and Computation2993-52102025-03-0121n/an/a10.1029/2024JH000402Decoding the Spatiotemporal Complexities of the Permafrost Carbon Feedback With Multimodal Ensemble LearningB. A. Gay0N. J. Pastick1J. D. Watts2A. H. Armstrong3K. R. Miner4C. E. Miller5Jet Propulsion Laboratory California Institute of Technology Pasadena CA USAUnited States Geological Survey Earth Resources Observation and Science Center Sioux Falls SD USAMontana State University Bozeman MT USANASA Goddard Space Flight Center Greenbelt MD USAJet Propulsion Laboratory California Institute of Technology Pasadena CA USAJet Propulsion Laboratory California Institute of Technology Pasadena CA USAAbstract Complex nonlinear relationships exist between the permafrost thermal state, active layer thickness, and terrestrial carbon cycle dynamics. In Arctic and boreal Alaska, significant uncertainties characterize the spatiotemporal rate and magnitude of permafrost degradation and the permafrost carbon feedback, with increasing recognition of the importance of thawing mechanisms. The challenges of monitoring sub‐surface phenomena with remote sensing technology further complicate the issue. There is an urgent need to understand how and to what extent thawing permafrost destabilizes the carbon balance in Alaska and to characterize the feedback involved. In this research, we use our artificial intelligence‐driven model GeoCryoAI to quantify permafrost carbon dynamics in Alaska. The GeoCryoAI model uses a hybridized process‐constrained ensemble learning framework to simultaneously ingest, scale, and analyze in situ measurements, remote sensing observations, and process‐based modeling outputs with disparate spatiotemporal sampling and data densities. We evaluated prior naïve (a) persistence and (b) teacher forcing approaches relative to (c) time‐delayed GeoCryoAI simulations, yielding the following error metrics (RMSE) for active layer thickness (ALT), methane (CH4), and carbon dioxide (CO2), respectively: 1.997, 1.327, 1.007 cm [1963–2022]; 0.884, 0.715, 0.694 nmol CH4km−2 month−1 [1994–2022]; 1.906, 0.697, 0.213 µmol CO2km−2 month−1 [1994–2022]. Our approach overcomes traditional model inefficiencies and resolves spatiotemporal disparities. GeoCryoAI captures abrupt and persistent changes while introducing a novel methodology for assimilating contemporaneous information at various scales. We describe GeoCryoAI, the methodology, our results, and plans for future applications.https://doi.org/10.1029/2024JH000402permafrost carbon feedbackcryosphereartificial intelligenceremote sensingclimate changecarbon cycling, modeling |
| spellingShingle | B. A. Gay N. J. Pastick J. D. Watts A. H. Armstrong K. R. Miner C. E. Miller Decoding the Spatiotemporal Complexities of the Permafrost Carbon Feedback With Multimodal Ensemble Learning Journal of Geophysical Research: Machine Learning and Computation permafrost carbon feedback cryosphere artificial intelligence remote sensing climate change carbon cycling, modeling |
| title | Decoding the Spatiotemporal Complexities of the Permafrost Carbon Feedback With Multimodal Ensemble Learning |
| title_full | Decoding the Spatiotemporal Complexities of the Permafrost Carbon Feedback With Multimodal Ensemble Learning |
| title_fullStr | Decoding the Spatiotemporal Complexities of the Permafrost Carbon Feedback With Multimodal Ensemble Learning |
| title_full_unstemmed | Decoding the Spatiotemporal Complexities of the Permafrost Carbon Feedback With Multimodal Ensemble Learning |
| title_short | Decoding the Spatiotemporal Complexities of the Permafrost Carbon Feedback With Multimodal Ensemble Learning |
| title_sort | decoding the spatiotemporal complexities of the permafrost carbon feedback with multimodal ensemble learning |
| topic | permafrost carbon feedback cryosphere artificial intelligence remote sensing climate change carbon cycling, modeling |
| url | https://doi.org/10.1029/2024JH000402 |
| work_keys_str_mv | AT bagay decodingthespatiotemporalcomplexitiesofthepermafrostcarbonfeedbackwithmultimodalensemblelearning AT njpastick decodingthespatiotemporalcomplexitiesofthepermafrostcarbonfeedbackwithmultimodalensemblelearning AT jdwatts decodingthespatiotemporalcomplexitiesofthepermafrostcarbonfeedbackwithmultimodalensemblelearning AT aharmstrong decodingthespatiotemporalcomplexitiesofthepermafrostcarbonfeedbackwithmultimodalensemblelearning AT krminer decodingthespatiotemporalcomplexitiesofthepermafrostcarbonfeedbackwithmultimodalensemblelearning AT cemiller decodingthespatiotemporalcomplexitiesofthepermafrostcarbonfeedbackwithmultimodalensemblelearning |