A Deep Learning Parameterization for Ozone Dry Deposition Velocities
The loss of ozone to terrestrial and aquatic systems, known as dry deposition, is a highly uncertain process governed by turbulent transport, interfacial chemistry, and plant physiology. We demonstrate the value of using Deep Neural Networks (DNN) in predicting ozone dry deposition velocities. We fi...
| Main Authors: | , , , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | English |
| Published: |
American Geophysical Union (AGU)
2020
|
| Online Access: | https://hdl.handle.net/1721.1/125517 |
| _version_ | 1826192641278083072 |
|---|---|
| author | Silva, Sam James Heald, Colette L. Ravela, Sai Mammarella, I. Munger, J. W. |
| author2 | Massachusetts Institute of Technology. Department of Civil and Environmental Engineering |
| author_facet | Massachusetts Institute of Technology. Department of Civil and Environmental Engineering Silva, Sam James Heald, Colette L. Ravela, Sai Mammarella, I. Munger, J. W. |
| author_sort | Silva, Sam James |
| collection | MIT |
| description | The loss of ozone to terrestrial and aquatic systems, known as dry deposition, is a highly uncertain process governed by turbulent transport, interfacial chemistry, and plant physiology. We demonstrate the value of using Deep Neural Networks (DNN) in predicting ozone dry deposition velocities. We find that a feedforward DNN trained on observations from a coniferous forest site (Hyytiälä, Finland) can predict hourly ozone dry deposition velocities at a mixed forest site (Harvard Forest, Massachusetts) more accurately than modern theoretical models, with a reduction in the normalized mean bias (0.05 versus ~0.1). The same DNN model, when driven by assimilated meteorology at 2° × 2.5° spatial resolution, outperforms the Wesely scheme as implemented in the GEOS-Chem model. With more available training data from other climate and ecological zones, this methodology could yield a generalizable DNN suitable for global models. |
| first_indexed | 2024-09-23T09:26:58Z |
| format | Article |
| id | mit-1721.1/125517 |
| institution | Massachusetts Institute of Technology |
| language | English |
| last_indexed | 2024-09-23T09:26:58Z |
| publishDate | 2020 |
| publisher | American Geophysical Union (AGU) |
| record_format | dspace |
| spelling | mit-1721.1/1255172022-09-30T14:27:35Z A Deep Learning Parameterization for Ozone Dry Deposition Velocities Silva, Sam James Heald, Colette L. Ravela, Sai Mammarella, I. Munger, J. W. Massachusetts Institute of Technology. Department of Civil and Environmental Engineering Massachusetts Institute of Technology. Department of Earth, Atmospheric, and Planetary Sciences The loss of ozone to terrestrial and aquatic systems, known as dry deposition, is a highly uncertain process governed by turbulent transport, interfacial chemistry, and plant physiology. We demonstrate the value of using Deep Neural Networks (DNN) in predicting ozone dry deposition velocities. We find that a feedforward DNN trained on observations from a coniferous forest site (Hyytiälä, Finland) can predict hourly ozone dry deposition velocities at a mixed forest site (Harvard Forest, Massachusetts) more accurately than modern theoretical models, with a reduction in the normalized mean bias (0.05 versus ~0.1). The same DNN model, when driven by assimilated meteorology at 2° × 2.5° spatial resolution, outperforms the Wesely scheme as implemented in the GEOS-Chem model. With more available training data from other climate and ecological zones, this methodology could yield a generalizable DNN suitable for global models. NSF (grant 1564495) U.S. DOE Office of Science (contract DE-AC02-05CH11231) NSF (DEB-1237491) 2020-05-27T19:02:10Z 2020-05-27T19:02:10Z 2019-01 2018-10 2020-05-27T17:37:29Z Article http://purl.org/eprint/type/JournalArticle 1944-8007 0094-8276 https://hdl.handle.net/1721.1/125517 Silva, S. J., Heald, C. L., Ravela, S., Mammarella, I., & Munger, J. W. (2019). A deep learning parameterization for ozone dry deposition velocities. Geophysical Research Letters, 46, 983–989. https://doi.org/10.1029/2018GL081049 en 10.1029/2018GL081049 Geophysical Research Letters Article is made available in accordance with the publisher's policy and may be subject to US copyright law. Please refer to the publisher's site for terms of use. application/pdf American Geophysical Union (AGU) MIT web domain |
| spellingShingle | Silva, Sam James Heald, Colette L. Ravela, Sai Mammarella, I. Munger, J. W. A Deep Learning Parameterization for Ozone Dry Deposition Velocities |
| title | A Deep Learning Parameterization for Ozone Dry Deposition Velocities |
| title_full | A Deep Learning Parameterization for Ozone Dry Deposition Velocities |
| title_fullStr | A Deep Learning Parameterization for Ozone Dry Deposition Velocities |
| title_full_unstemmed | A Deep Learning Parameterization for Ozone Dry Deposition Velocities |
| title_short | A Deep Learning Parameterization for Ozone Dry Deposition Velocities |
| title_sort | deep learning parameterization for ozone dry deposition velocities |
| url | https://hdl.handle.net/1721.1/125517 |
| work_keys_str_mv | AT silvasamjames adeeplearningparameterizationforozonedrydepositionvelocities AT healdcolettel adeeplearningparameterizationforozonedrydepositionvelocities AT ravelasai adeeplearningparameterizationforozonedrydepositionvelocities AT mammarellai adeeplearningparameterizationforozonedrydepositionvelocities AT mungerjw adeeplearningparameterizationforozonedrydepositionvelocities AT silvasamjames deeplearningparameterizationforozonedrydepositionvelocities AT healdcolettel deeplearningparameterizationforozonedrydepositionvelocities AT ravelasai deeplearningparameterizationforozonedrydepositionvelocities AT mammarellai deeplearningparameterizationforozonedrydepositionvelocities AT mungerjw deeplearningparameterizationforozonedrydepositionvelocities |