New particle formation event detection with Mask R-CNN
<p>Atmospheric new particle formation (NPF) is an important source of climate-relevant aerosol particles which has been observed at many locations globally. To study this phenomenon, the first step is to identify whether an NPF event occurs or not on a given day. In practice, NPF event identif...
| Main Authors: | , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Copernicus Publications
2022-01-01
|
| Series: | Atmospheric Chemistry and Physics |
| Online Access: | https://acp.copernicus.org/articles/22/1293/2022/acp-22-1293-2022.pdf |
| _version_ | 1819243247739863040 |
|---|---|
| author | P. Su P. Su J. Joutsensaari L. Dada L. Dada M. A. Zaidan M. A. Zaidan T. Nieminen T. Nieminen X. Li Y. Wu S. Decesari S. Tarkoma T. Petäjä T. Petäjä M. Kulmala M. Kulmala P. Pellikka P. Pellikka |
| author_facet | P. Su P. Su J. Joutsensaari L. Dada L. Dada M. A. Zaidan M. A. Zaidan T. Nieminen T. Nieminen X. Li Y. Wu S. Decesari S. Tarkoma T. Petäjä T. Petäjä M. Kulmala M. Kulmala P. Pellikka P. Pellikka |
| author_sort | P. Su |
| collection | DOAJ |
| description | <p>Atmospheric new particle formation (NPF) is an important source of climate-relevant aerosol particles which has been observed at many locations globally. To study this phenomenon, the first step is to identify whether an NPF event occurs or not on a given day. In practice, NPF event identification is performed visually by classifying the NPF event or non-event days from the particle number size distribution surface plots. Unfortunately, this day-by-day visual classification is time-consuming and labor-intensive, and the identification process renders subjective results. To detect NPF events automatically, we regard the visual signature (banana shape) which has been observed all over the world in NPF surface plots as a special kind of object, and a deep learning model called Mask R-CNN is applied to localize the spatial layouts of NPF events in their surface plots. Utilizing only 358 human-annotated masks on data from the Station for Measuring Ecosystem–Atmosphere Relations (SMEAR) II station (Hyytiälä, Finland), the Mask R-CNN model was successfully generalized for three SMEAR stations in Finland and the San Pietro Capofiume (SPC) station in Italy. In addition to the detection of NPF events (especially the strongest events), the presented method can determine the growth rates, start times, and end times for NPF events automatically. The automatically determined growth rates agree with the manually determined growth rates. The statistical results validate the potential of applying the proposed method to different sites, which will improve the automatic level for NPF event detection and analysis. Furthermore, the proposed automatic NPF event analysis method can minimize subjectivity compared with human-made analysis, especially when long-term data series are analyzed and statistical comparisons between different sites are needed for event characteristics such as the start and end times, thereby saving time and effort for scientists studying NPF events.</p> |
| first_indexed | 2024-12-23T14:52:40Z |
| format | Article |
| id | doaj.art-d1ef5b99ff16409daca9e4e537722127 |
| institution | Directory Open Access Journal |
| issn | 1680-7316 1680-7324 |
| language | English |
| last_indexed | 2024-12-23T14:52:40Z |
| publishDate | 2022-01-01 |
| publisher | Copernicus Publications |
| record_format | Article |
| series | Atmospheric Chemistry and Physics |
| spelling | doaj.art-d1ef5b99ff16409daca9e4e5377221272022-12-21T17:42:54ZengCopernicus PublicationsAtmospheric Chemistry and Physics1680-73161680-73242022-01-01221293130910.5194/acp-22-1293-2022New particle formation event detection with Mask R-CNNP. Su0P. Su1J. Joutsensaari2L. Dada3L. Dada4M. A. Zaidan5M. A. Zaidan6T. Nieminen7T. Nieminen8X. Li9Y. Wu10S. Decesari11S. Tarkoma12T. Petäjä13T. Petäjä14M. Kulmala15M. Kulmala16P. Pellikka17P. Pellikka18Department of Geosciences and Geography, University of Helsinki, 00014 Helsinki, FinlandInstitute for Atmospheric and Earth System Research (INAR/Physics), Faculty of Science, University of Helsinki, 00014 Helsinki, FinlandDepartment of Applied Physics, University of Eastern Finland, P.O. Box 1627, 70211 Kuopio, FinlandExtreme Environments Research Laboratory, École Polytechnique Fédérale de Lausanne (EPFL) Valais, 1951 Sion, SwitzerlandLaboratory of Atmospheric Chemistry, Paul Scherrer Institute, 5232 Villigen, SwitzerlandInstitute for Atmospheric and Earth System Research (INAR/Physics), Faculty of Science, University of Helsinki, 00014 Helsinki, FinlandJoint International Research Laboratory of Atmospheric and Earth System Sciences, School of Atmospheric Sciences, Nanjing University, Nanjing, 210023, ChinaInstitute for Atmospheric and Earth System Research (INAR/Physics), Faculty of Science, University of Helsinki, 00014 Helsinki, FinlandInstitute for Atmospheric and Earth System Research (INAR/Forest Sciences), Faculty of Agriculture and Forestry, University of Helsinki, 00014 Helsinki, FinlandInstitute for Atmospheric and Earth System Research (INAR/Physics), Faculty of Science, University of Helsinki, 00014 Helsinki, FinlandInstitute for Atmospheric and Earth System Research (INAR/Physics), Faculty of Science, University of Helsinki, 00014 Helsinki, FinlandInstitute of Atmospheric and Climate Sciences, National Research Council of Italy (CNR), 40129, Bologna, ItalyDepartment of Computer Science, Faculty of Science, University of Helsinki, 00014 Helsinki, FinlandInstitute for Atmospheric and Earth System Research (INAR/Physics), Faculty of Science, University of Helsinki, 00014 Helsinki, FinlandJoint International Research Laboratory of Atmospheric and Earth System Sciences, School of Atmospheric Sciences, Nanjing University, Nanjing, 210023, ChinaInstitute for Atmospheric and Earth System Research (INAR/Physics), Faculty of Science, University of Helsinki, 00014 Helsinki, FinlandJoint International Research Laboratory of Atmospheric and Earth System Sciences, School of Atmospheric Sciences, Nanjing University, Nanjing, 210023, ChinaDepartment of Geosciences and Geography, University of Helsinki, 00014 Helsinki, FinlandInstitute for Atmospheric and Earth System Research (INAR/Physics), Faculty of Science, University of Helsinki, 00014 Helsinki, Finland<p>Atmospheric new particle formation (NPF) is an important source of climate-relevant aerosol particles which has been observed at many locations globally. To study this phenomenon, the first step is to identify whether an NPF event occurs or not on a given day. In practice, NPF event identification is performed visually by classifying the NPF event or non-event days from the particle number size distribution surface plots. Unfortunately, this day-by-day visual classification is time-consuming and labor-intensive, and the identification process renders subjective results. To detect NPF events automatically, we regard the visual signature (banana shape) which has been observed all over the world in NPF surface plots as a special kind of object, and a deep learning model called Mask R-CNN is applied to localize the spatial layouts of NPF events in their surface plots. Utilizing only 358 human-annotated masks on data from the Station for Measuring Ecosystem–Atmosphere Relations (SMEAR) II station (Hyytiälä, Finland), the Mask R-CNN model was successfully generalized for three SMEAR stations in Finland and the San Pietro Capofiume (SPC) station in Italy. In addition to the detection of NPF events (especially the strongest events), the presented method can determine the growth rates, start times, and end times for NPF events automatically. The automatically determined growth rates agree with the manually determined growth rates. The statistical results validate the potential of applying the proposed method to different sites, which will improve the automatic level for NPF event detection and analysis. Furthermore, the proposed automatic NPF event analysis method can minimize subjectivity compared with human-made analysis, especially when long-term data series are analyzed and statistical comparisons between different sites are needed for event characteristics such as the start and end times, thereby saving time and effort for scientists studying NPF events.</p>https://acp.copernicus.org/articles/22/1293/2022/acp-22-1293-2022.pdf |
| spellingShingle | P. Su P. Su J. Joutsensaari L. Dada L. Dada M. A. Zaidan M. A. Zaidan T. Nieminen T. Nieminen X. Li Y. Wu S. Decesari S. Tarkoma T. Petäjä T. Petäjä M. Kulmala M. Kulmala P. Pellikka P. Pellikka New particle formation event detection with Mask R-CNN Atmospheric Chemistry and Physics |
| title | New particle formation event detection with Mask R-CNN |
| title_full | New particle formation event detection with Mask R-CNN |
| title_fullStr | New particle formation event detection with Mask R-CNN |
| title_full_unstemmed | New particle formation event detection with Mask R-CNN |
| title_short | New particle formation event detection with Mask R-CNN |
| title_sort | new particle formation event detection with mask r cnn |
| url | https://acp.copernicus.org/articles/22/1293/2022/acp-22-1293-2022.pdf |
| work_keys_str_mv | AT psu newparticleformationeventdetectionwithmaskrcnn AT psu newparticleformationeventdetectionwithmaskrcnn AT jjoutsensaari newparticleformationeventdetectionwithmaskrcnn AT ldada newparticleformationeventdetectionwithmaskrcnn AT ldada newparticleformationeventdetectionwithmaskrcnn AT mazaidan newparticleformationeventdetectionwithmaskrcnn AT mazaidan newparticleformationeventdetectionwithmaskrcnn AT tnieminen newparticleformationeventdetectionwithmaskrcnn AT tnieminen newparticleformationeventdetectionwithmaskrcnn AT xli newparticleformationeventdetectionwithmaskrcnn AT ywu newparticleformationeventdetectionwithmaskrcnn AT sdecesari newparticleformationeventdetectionwithmaskrcnn AT starkoma newparticleformationeventdetectionwithmaskrcnn AT tpetaja newparticleformationeventdetectionwithmaskrcnn AT tpetaja newparticleformationeventdetectionwithmaskrcnn AT mkulmala newparticleformationeventdetectionwithmaskrcnn AT mkulmala newparticleformationeventdetectionwithmaskrcnn AT ppellikka newparticleformationeventdetectionwithmaskrcnn AT ppellikka newparticleformationeventdetectionwithmaskrcnn |