Exploring Machine Learning to Correct Satellite-Derived Sea Surface Temperatures
Machine learning techniques are attractive tools to establish statistical models with a high degree of non linearity. They require a large amount of data to be trained and are therefore particularly suited to analysing remote sensing data. This work is an attempt at using advanced statistical method...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2018-02-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | http://www.mdpi.com/2072-4292/10/2/224 |
| _version_ | 1818870363661008896 |
|---|---|
| author | Stéphane Saux Picart Pierre Tandeo Emmanuelle Autret Blandine Gausset |
| author_facet | Stéphane Saux Picart Pierre Tandeo Emmanuelle Autret Blandine Gausset |
| author_sort | Stéphane Saux Picart |
| collection | DOAJ |
| description | Machine learning techniques are attractive tools to establish statistical models with a high degree of non linearity. They require a large amount of data to be trained and are therefore particularly suited to analysing remote sensing data. This work is an attempt at using advanced statistical methods of machine learning to predict the bias between Sea Surface Temperature (SST) derived from infrared remote sensing and ground “truth” from drifting buoy measurements. A large dataset of collocation between satellite SST and in situ SST is explored. Four regression models are used: Simple multi-linear regression, Least Square Shrinkage and Selection Operator (LASSO), Generalised Additive Model (GAM) and random forest. In the case of geostationary satellites for which a large number of collocations is available, results show that the random forest model is the best model to predict the systematic errors and it is computationally fast, making it a good candidate for operational processing. It is able to explain nearly 31% of the total variance of the bias (in comparison to about 24% for the multi-linear regression model). |
| first_indexed | 2024-12-19T12:05:50Z |
| format | Article |
| id | doaj.art-fb6e1292325a475bbc197527bd2d5051 |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-12-19T12:05:50Z |
| publishDate | 2018-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-fb6e1292325a475bbc197527bd2d50512022-12-21T20:22:22ZengMDPI AGRemote Sensing2072-42922018-02-0110222410.3390/rs10020224rs10020224Exploring Machine Learning to Correct Satellite-Derived Sea Surface TemperaturesStéphane Saux Picart0Pierre Tandeo1Emmanuelle Autret2Blandine Gausset3Météo-France/Centre de Météorologie Spatiale, Avenue de Lorraine, B.P. 50747, 22307 Lannion CEDEX, FranceIMT Atlantique, Lab-STICC, UBL, 29238 Brest, FranceIfremer, Laboratoire d’Océanographie Physique et Spatiale, ZI Pointe du Diable CS 10070, 29280 Plouzané, FranceMétéo-France/Centre de Météorologie Spatiale, Avenue de Lorraine, B.P. 50747, 22307 Lannion CEDEX, FranceMachine learning techniques are attractive tools to establish statistical models with a high degree of non linearity. They require a large amount of data to be trained and are therefore particularly suited to analysing remote sensing data. This work is an attempt at using advanced statistical methods of machine learning to predict the bias between Sea Surface Temperature (SST) derived from infrared remote sensing and ground “truth” from drifting buoy measurements. A large dataset of collocation between satellite SST and in situ SST is explored. Four regression models are used: Simple multi-linear regression, Least Square Shrinkage and Selection Operator (LASSO), Generalised Additive Model (GAM) and random forest. In the case of geostationary satellites for which a large number of collocations is available, results show that the random forest model is the best model to predict the systematic errors and it is computationally fast, making it a good candidate for operational processing. It is able to explain nearly 31% of the total variance of the bias (in comparison to about 24% for the multi-linear regression model).http://www.mdpi.com/2072-4292/10/2/224machine learningsystematic errorsea surface temperaturerandom forest |
| spellingShingle | Stéphane Saux Picart Pierre Tandeo Emmanuelle Autret Blandine Gausset Exploring Machine Learning to Correct Satellite-Derived Sea Surface Temperatures Remote Sensing machine learning systematic error sea surface temperature random forest |
| title | Exploring Machine Learning to Correct Satellite-Derived Sea Surface Temperatures |
| title_full | Exploring Machine Learning to Correct Satellite-Derived Sea Surface Temperatures |
| title_fullStr | Exploring Machine Learning to Correct Satellite-Derived Sea Surface Temperatures |
| title_full_unstemmed | Exploring Machine Learning to Correct Satellite-Derived Sea Surface Temperatures |
| title_short | Exploring Machine Learning to Correct Satellite-Derived Sea Surface Temperatures |
| title_sort | exploring machine learning to correct satellite derived sea surface temperatures |
| topic | machine learning systematic error sea surface temperature random forest |
| url | http://www.mdpi.com/2072-4292/10/2/224 |
| work_keys_str_mv | AT stephanesauxpicart exploringmachinelearningtocorrectsatellitederivedseasurfacetemperatures AT pierretandeo exploringmachinelearningtocorrectsatellitederivedseasurfacetemperatures AT emmanuelleautret exploringmachinelearningtocorrectsatellitederivedseasurfacetemperatures AT blandinegausset exploringmachinelearningtocorrectsatellitederivedseasurfacetemperatures |