Expanding Mars’s Climate Modeling: Interpretable Machine Learning for Modeling Mars Science Laboratory Relative Humidity
For the past several decades, numerous attempts have been made to model the climate of Mars, with extensive studies focusing on the planet’s dynamics and climate. While physical modeling and data assimilation approaches have made significant progress, uncertainties persist in comprehensively capturi...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IOP Publishing
2024-01-01
|
| Series: | The Planetary Science Journal |
| Subjects: | |
| Online Access: | https://doi.org/10.3847/PSJ/ad25fd |
| _version_ | 1797231838092066816 |
|---|---|
| author | Nour Abdelmoneim Dattaraj B. Dhuri Dimitra Atri Germán Martínez |
| author_facet | Nour Abdelmoneim Dattaraj B. Dhuri Dimitra Atri Germán Martínez |
| author_sort | Nour Abdelmoneim |
| collection | DOAJ |
| description | For the past several decades, numerous attempts have been made to model the climate of Mars, with extensive studies focusing on the planet’s dynamics and climate. While physical modeling and data assimilation approaches have made significant progress, uncertainties persist in comprehensively capturing the complexities of the Martian climate. We propose a novel approach to Martian climate modeling by leveraging machine-learning techniques that have shown remarkable success in Earth climate modeling. Our study presents a deep neural network designed to model relative humidity in Gale crater, as measured by NASA’s Mars Science Laboratory “Curiosity” rover. By utilizing meteorological variables produced by the Mars Planetary Climate Model, our model accurately predicts relative humidity with a mean error of 3% and an R ^2 score of 0.92 over the range of relative humidity compared. Furthermore, we present an approach to predict quantile ranges of relative humidity, catering to applications that require a range of values. To address the challenge of interpretability associated with machine-learning models, we utilize an interpretable model architecture and conduct an in-depth analysis of its decision-making processes. We find that our neural network can model relative humidity at Gale crater using a few meteorological variables, with the monthly mean surface H _2 O layer, planetary boundary layer height, convective wind speed, and solar zenith angle being the primary contributors. In addition to providing an efficient method for modeling climate variables on Mars, this approach can also be utilized to expand on current data sets by filling spatial and temporal gaps in observations. |
| first_indexed | 2024-04-24T15:50:44Z |
| format | Article |
| id | doaj.art-dcaeaebfb44a4b9a8012c49188cb89e7 |
| institution | Directory Open Access Journal |
| issn | 2632-3338 |
| language | English |
| last_indexed | 2024-04-24T15:50:44Z |
| publishDate | 2024-01-01 |
| publisher | IOP Publishing |
| record_format | Article |
| series | The Planetary Science Journal |
| spelling | doaj.art-dcaeaebfb44a4b9a8012c49188cb89e72024-04-01T12:34:53ZengIOP PublishingThe Planetary Science Journal2632-33382024-01-01548610.3847/PSJ/ad25fdExpanding Mars’s Climate Modeling: Interpretable Machine Learning for Modeling Mars Science Laboratory Relative HumidityNour Abdelmoneim0https://orcid.org/0009-0006-7197-6408Dattaraj B. Dhuri1Dimitra Atri2Germán Martínez3Center for Astrophysics and Space Science, New York University Abu Dhabi, P.O. Box 129188 , Saadiyat Island, Abu Dhabi, UAE ; atri@nyu.eduCenter for Astrophysics and Space Science, New York University Abu Dhabi, P.O. Box 129188 , Saadiyat Island, Abu Dhabi, UAE ; atri@nyu.eduCenter for Astrophysics and Space Science, New York University Abu Dhabi, P.O. Box 129188 , Saadiyat Island, Abu Dhabi, UAE ; atri@nyu.eduLunar and Planetary Institute, Universities Space Research Association, Houston , TX 77058, USAFor the past several decades, numerous attempts have been made to model the climate of Mars, with extensive studies focusing on the planet’s dynamics and climate. While physical modeling and data assimilation approaches have made significant progress, uncertainties persist in comprehensively capturing the complexities of the Martian climate. We propose a novel approach to Martian climate modeling by leveraging machine-learning techniques that have shown remarkable success in Earth climate modeling. Our study presents a deep neural network designed to model relative humidity in Gale crater, as measured by NASA’s Mars Science Laboratory “Curiosity” rover. By utilizing meteorological variables produced by the Mars Planetary Climate Model, our model accurately predicts relative humidity with a mean error of 3% and an R ^2 score of 0.92 over the range of relative humidity compared. Furthermore, we present an approach to predict quantile ranges of relative humidity, catering to applications that require a range of values. To address the challenge of interpretability associated with machine-learning models, we utilize an interpretable model architecture and conduct an in-depth analysis of its decision-making processes. We find that our neural network can model relative humidity at Gale crater using a few meteorological variables, with the monthly mean surface H _2 O layer, planetary boundary layer height, convective wind speed, and solar zenith angle being the primary contributors. In addition to providing an efficient method for modeling climate variables on Mars, this approach can also be utilized to expand on current data sets by filling spatial and temporal gaps in observations.https://doi.org/10.3847/PSJ/ad25fdMarsPlanetary climatesHumidityNeural networks |
| spellingShingle | Nour Abdelmoneim Dattaraj B. Dhuri Dimitra Atri Germán Martínez Expanding Mars’s Climate Modeling: Interpretable Machine Learning for Modeling Mars Science Laboratory Relative Humidity The Planetary Science Journal Mars Planetary climates Humidity Neural networks |
| title | Expanding Mars’s Climate Modeling: Interpretable Machine Learning for Modeling Mars Science Laboratory Relative Humidity |
| title_full | Expanding Mars’s Climate Modeling: Interpretable Machine Learning for Modeling Mars Science Laboratory Relative Humidity |
| title_fullStr | Expanding Mars’s Climate Modeling: Interpretable Machine Learning for Modeling Mars Science Laboratory Relative Humidity |
| title_full_unstemmed | Expanding Mars’s Climate Modeling: Interpretable Machine Learning for Modeling Mars Science Laboratory Relative Humidity |
| title_short | Expanding Mars’s Climate Modeling: Interpretable Machine Learning for Modeling Mars Science Laboratory Relative Humidity |
| title_sort | expanding mars s climate modeling interpretable machine learning for modeling mars science laboratory relative humidity |
| topic | Mars Planetary climates Humidity Neural networks |
| url | https://doi.org/10.3847/PSJ/ad25fd |
| work_keys_str_mv | AT nourabdelmoneim expandingmarssclimatemodelinginterpretablemachinelearningformodelingmarssciencelaboratoryrelativehumidity AT dattarajbdhuri expandingmarssclimatemodelinginterpretablemachinelearningformodelingmarssciencelaboratoryrelativehumidity AT dimitraatri expandingmarssclimatemodelinginterpretablemachinelearningformodelingmarssciencelaboratoryrelativehumidity AT germanmartinez expandingmarssclimatemodelinginterpretablemachinelearningformodelingmarssciencelaboratoryrelativehumidity |