Data-driven cloud cover parameterizations for the ICON earth system model using deep learning and symbolic regression
This thesis delves into the improvement of cloud parameterizations in climate models through machine learning trained on coarse-grained output from high-resolution simulations. Utilizing the ICOsahedral Non-hydrostatic (ICON) modeling framework, it specifically targets the enhancement of cloud cover...
Ausführliche Beschreibung
Gespeichert in:
| Autor*in: |
Grundner, Arthur [verfasserIn] Eyring, Veronika [akademischer betreuerIn] Gentine, Pierre [akademischer betreuerIn] |
|---|---|
| Körperschaften: |
Universität Bremen [Grad-verleihende Institution] |
| Hochschulschrift: |
Dissertation ; Universität Bremen ; 2024 |
| Format: |
E-Book |
|---|---|
| Sprache: |
Englisch |
| Erschienen: |
Bremen: 2024 |
|---|
| Rechteinformationen: |
Open Access Namensnennung 4.0 International ; CC BY 4.0 |
|---|
| Schlagwörter: | |
|---|---|
| Formangabe: |
Hochschulschrift |
| Umfang: |
1 Online-Ressource (xi, 140 Seiten) ; Illustrationen, Diagramme |
|---|
| Weitere Ausgabe: |
Erscheint auch als Druck-Ausgabe Grundner, Arthur: Data-driven cloud cover parameterizations for the ICON earth system model using deep learning and symbolic regression - Bremen, 2023 |
|---|
| Links: |
Link aufrufen |
|---|
| DOI / URN: |
urn:nbn:de:gbv:46-elib77397 10.26092/elib/2821 |
|---|
| Katalog-ID: |
1883908477 |
|---|
| LEADER | 01000cam a2200265 4500 | ||
|---|---|---|---|
| 001 | 1883908477 | ||
| 003 | DE-627 | ||
| 005 | 20240912032145.0 | ||
| 007 | cr uuu---uuuuu | ||
| 008 | 240320s2024 gw |||||om 00| ||eng c | ||
| 024 | 7 | |a urn:nbn:de:gbv:46-elib77397 |2 urn | |
| 024 | 7 | |a 10.26092/elib/2821 |2 doi | |
| 035 | |a (DE-627)1883908477 | ||
| 035 | |a (DE-599)KXP1883908477 | ||
| 035 | |a (OCoLC)1427339102 | ||
| 035 | |a (OAEPFHB)elib/2821 | ||
| 040 | |a DE-627 |b ger |c DE-627 |e rda | ||
| 041 | |a eng | ||
| 044 | |c XA-DE-HB | ||
| 082 | 0 | 4 | |a 551.576 |q DE-101 |
| 082 | 0 | 4 | |a 550 |q DE-101 |
| 100 | 1 | |a Grundner, Arthur |e verfasserin |0 (orcid)0000-0002-3765-242X |4 aut | |
| 245 | 1 | 0 | |a Data-driven cloud cover parameterizations for the ICON earth system model using deep learning and symbolic regression |c Arthur Grundner |
| 264 | 1 | |a Bremen |c 2024 | |
| 300 | |a 1 Online-Ressource (xi, 140 Seiten) |b Illustrationen, Diagramme | ||
| 336 | |a Text |b txt |2 rdacontent | ||
| 337 | |a Computermedien |b c |2 rdamedia | ||
| 338 | |a Online-Ressource |b cr |2 rdacarrier | ||
| 502 | |b Dissertation |c Universität Bremen |d 2024 | ||
| 506 | 0 | |q DE-46 |a Open Access |e Controlled Vocabulary for Access Rights |u http://purl.org/coar/access_right/c_abf2 | |
| 520 | |a This thesis delves into the improvement of cloud parameterizations in climate models through machine learning trained on coarse-grained output from high-resolution simulations. Utilizing the ICOsahedral Non-hydrostatic (ICON) modeling framework, it specifically targets the enhancement of cloud cover parameterization within the ICON Earth System Model. Three types of neural networks (NNs) differing in vertical locality are developed to estimate cloud cover, with globally trained NNs even applicable to distinct regional simulations. Interpretability analysis exposes model-specific biases and local relationships with the thermodynamic environment. Despite achieving high predictive performance, NNs necessitate post-hoc interpretation tools. To tackle this issue, a combined hierarchical modeling framework incorporating symbolic regression, feature selection, and physical constraints is proposed. The resulting equations, characterized by simplicity and physical consistency, attain performance comparable to NNs while demonstrating superior transferability to other realistic datasets. Our best equation adeptly captures cloud cover distributions across various regimes, notably excelling in representing marine stratocumulus clouds by learning to utilize the vertical relative humidity gradient. This research underscores the potential of deep learning in achieving accurate cloud parameterizations and emphasizes the effective role of symbolic regression in deriving interpretable, consistent equations for cloud cover. | ||
| 540 | |q DE-46 |a Namensnennung 4.0 International |f CC BY 4.0 |2 cc |u https://creativecommons.org/licenses/by/4.0/ | ||
| 583 | 1 | |a Archivierung/Langzeitarchivierung gewährleistet |f PEHB |x XA-DE-HB |2 pdager |5 DE-46 | |
| 650 | 4 | |a Cloud Cover | |
| 650 | 4 | |a Parameterization | |
| 650 | 4 | |a Deep Learning | |
| 650 | 4 | |a Neural Networks | |
| 650 | 4 | |a ICON | |
| 650 | 4 | |a Pareto Frontier | |
| 650 | 4 | |a PySR | |
| 650 | 4 | |a Sequential Feature Selection | |
| 650 | 4 | |a Equation Discovery | |
| 650 | 4 | |a Physical Constraints | |
| 650 | 4 | |a Machine Learning | |
| 650 | 4 | |a Symbolic Regression | |
| 650 | 4 | |a SHAP | |
| 655 | 7 | |a Hochschulschrift |0 (DE-588)4113937-9 |0 (DE-627)105825778 |0 (DE-576)209480580 |2 gnd-content | |
| 700 | 1 | |a Eyring, Veronika |e akademischer betreuerin |0 (DE-588)1255681446 |0 (DE-627)1799770168 |4 dgs | |
| 700 | 1 | |a Gentine, Pierre |e akademischer betreuerin |4 dgs | |
| 710 | 2 | |a Universität Bremen |e Grad-verleihende Institution |0 (DE-588)2001386-3 |0 (DE-627)101380429 |0 (DE-576)191575038 |4 dgg | |
| 751 | |a Bremen |0 (DE-588)4008135-7 |0 (DE-627)106369636 |0 (DE-576)208874569 |4 uvp | ||
| 776 | 0 | 8 | |i Erscheint auch als |n Druck-Ausgabe |a Grundner, Arthur |t Data-driven cloud cover parameterizations for the ICON earth system model using deep learning and symbolic regression |d Bremen, 2023 |h xi, 140 Seiten |w (DE-627)1883908809 |
| 856 | 4 | 0 | |u https://doi.org/10.26092/elib/2821 |x Resolving-System |z kostenfrei |
| 856 | 4 | 0 | |u https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 |x Resolving-System |z kostenfrei |
| 856 | 4 | 0 | |u https://d-nb.info/1339510391/34 |x Langzeitarchivierung Nationalbibliothek |z kostenfrei |
| 856 | 4 | 0 | |u https://media.suub.uni-bremen.de/handle/elib/7739 |x Verlag |z kostenfrei |
| 912 | |a GBV-ODiss | ||
| 912 | |a GBV_ILN_20 | ||
| 912 | |a ISIL_DE-84 | ||
| 912 | |a SYSFLAG_1 | ||
| 912 | |a GBV_KXP | ||
| 912 | |a GBV_ILN_21 | ||
| 912 | |a ISIL_DE-46 | ||
| 912 | |a GBV_ILN_22 | ||
| 912 | |a ISIL_DE-18 | ||
| 912 | |a GBV_ILN_23 | ||
| 912 | |a ISIL_DE-830 | ||
| 912 | |a GBV_ILN_30 | ||
| 912 | |a ISIL_DE-104 | ||
| 912 | |a GBV_ILN_40 | ||
| 912 | |a ISIL_DE-7 | ||
| 912 | |a GBV_ILN_60 | ||
| 912 | |a ISIL_DE-705 | ||
| 912 | |a GBV_ILN_63 | ||
| 912 | |a ISIL_DE-Wim2 | ||
| 912 | |a GBV_ILN_70 | ||
| 912 | |a ISIL_DE-89 | ||
| 912 | |a GBV_ILN_105 | ||
| 912 | |a ISIL_DE-841 | ||
| 912 | |a GBV_ILN_110 | ||
| 912 | |a ISIL_DE-Luen4 | ||
| 912 | |a GBV_ILN_132 | ||
| 912 | |a ISIL_DE-959 | ||
| 912 | |a GBV_ILN_151 | ||
| 912 | |a ISIL_DE-546 | ||
| 912 | |a GBV_ILN_161 | ||
| 912 | |a ISIL_DE-960 | ||
| 912 | |a GBV_ILN_293 | ||
| 912 | |a ISIL_DE-960-3 | ||
| 912 | |a GBV_ILN_370 | ||
| 912 | |a ISIL_DE-1373 | ||
| 912 | |a GBV_ILN_2403 | ||
| 912 | |a ISIL_DE-LFER | ||
| 935 | |i DSpace | ||
| 951 | |a BO | ||
| 980 | |2 20 |1 01 |x 0084 |b 4593082714 |y x |z 12-10-24 | ||
| 980 | |2 21 |1 01 |x 0046 |b 4502027553 |h ebook_2024_dissbremen |k Kostenloser Zugriff |y zza |z 20-03-24 | ||
| 980 | |2 22 |1 01 |x 0018 |b 4593184053 |h SUBolrd |y xu |z 12-10-24 | ||
| 980 | |2 23 |1 01 |x 0830 |b 459323218X |h olr-d |y x |z 12-10-24 | ||
| 980 | |2 30 |1 01 |x 0104 |b 4593278708 |y z |z 12-10-24 | ||
| 980 | |2 40 |1 01 |x 0007 |b 4593314887 |y xsn |z 12-10-24 | ||
| 980 | |2 60 |1 01 |x 0705 |b 4593373239 |h OLRD |y z |z 12-10-24 | ||
| 980 | |2 63 |1 01 |x 3401 |b 4593426561 |h ORD |y x |z 12-10-24 | ||
| 980 | |2 70 |1 01 |x 0089 |b 4593481716 |y z |z 12-10-24 | ||
| 980 | |2 105 |1 01 |x 0841 |b 4593871697 |y z |z 12-10-24 | ||
| 980 | |2 110 |1 01 |x 3110 |b 4593582741 |y x |z 12-10-24 | ||
| 980 | |2 132 |1 01 |x 0959 |b 4593626374 |h OLR-DISS |y x |z 12-10-24 | ||
| 980 | |2 151 |1 01 |x 0546 |b 4593667224 |h OLR-ODISS |y z |z 12-10-24 | ||
| 980 | |2 161 |1 01 |x 0960 |b 4593691222 |h ORD |y z |z 12-10-24 | ||
| 980 | |2 293 |1 01 |x 3293 |b 4593820650 |h ORD |y z |z 12-10-24 | ||
| 980 | |2 370 |1 01 |x 4370 |b 4593854458 |y x |z 12-10-24 | ||
| 980 | |2 2403 |1 01 |x DE-LFER |b 4511167990 |c 00 |f --%%-- |d --%%-- |e n |j --%%-- |y l01 |z 12-04-24 | ||
| 981 | |2 20 |1 01 |x 0084 |r https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 | ||
| 981 | |2 21 |1 01 |x 0046 |r https://doi.org/10.26092/elib/2821 |z LF | ||
| 981 | |2 22 |1 01 |x 0018 |r https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 | ||
| 981 | |2 23 |1 01 |x 0830 |r https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 | ||
| 981 | |2 30 |1 01 |x 0104 |r https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 | ||
| 981 | |2 40 |1 01 |x 0007 |r https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 | ||
| 981 | |2 60 |1 01 |x 0705 |r https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 | ||
| 981 | |2 63 |1 01 |x 3401 |y E-Book |r https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 |z LF | ||
| 981 | |2 70 |1 01 |x 0089 |r https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 | ||
| 981 | |2 105 |1 01 |x 0841 |r https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 | ||
| 981 | |2 110 |1 01 |x 3110 |r https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 | ||
| 981 | |2 132 |1 01 |x 0959 |r https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 | ||
| 981 | |2 151 |1 01 |x 0546 |y Volltext |r https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 | ||
| 981 | |2 161 |1 01 |x 0960 |r https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 | ||
| 981 | |2 293 |1 01 |x 3293 |r https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 | ||
| 981 | |2 370 |1 01 |x 4370 |r https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 | ||
| 981 | |2 2403 |1 01 |x DE-LFER |r https://doi.org/10.26092/elib/2821 | ||
| 982 | |2 21 |1 00 |x DE-46 |8 00 |a Universität Bremen | ||
| 982 | |2 21 |1 00 |x DE-46 |8 00 |a Fachbereich 01: Physik/Elektrotechnik (FB 01) | ||
| 983 | |2 60 |1 01 |x 0705 |8 10 |a ho | ||
| 985 | |2 20 |1 01 |x 0084 |a OLRD | ||
| 985 | |2 110 |1 01 |x 3110 |a OLRD | ||
| 985 | |2 370 |1 01 |x 4370 |a OLRD | ||
| 995 | |2 21 |1 01 |x 0046 |a ebook_2024_dissbremen | ||
| 995 | |2 22 |1 01 |x 0018 |a SUBolrd | ||
| 995 | |2 23 |1 01 |x 0830 |a olr-d | ||
| 995 | |2 60 |1 01 |x 0705 |a OLRD | ||
| 995 | |2 63 |1 01 |x 3401 |a ORD | ||
| 995 | |2 132 |1 01 |x 0959 |a OLR-DISS | ||
| 995 | |2 151 |1 01 |x 0546 |a OLR-ODISS | ||
| 995 | |2 161 |1 01 |x 0960 |a ORD | ||
| 995 | |2 293 |1 01 |x 3293 |a ORD | ||
| 998 | |2 23 |1 01 |x 0830 |0 2024-10-12 10:31:16 | ||
| author_variant |
a g ag |
|---|---|
| matchkey_str |
grundnerarthureyringveronikagentinepierr:2024----:aarvnlucvraaeeiainfrhioerhytmoeuigepe |
| oclc_num |
1427339102 |
| hierarchy_sort_str |
2024 |
| publishDate |
2024 |
| allfields |
urn:nbn:de:gbv:46-elib77397 urn 10.26092/elib/2821 doi (DE-627)1883908477 (DE-599)KXP1883908477 (OCoLC)1427339102 (OAEPFHB)elib/2821 DE-627 ger DE-627 rda eng XA-DE-HB 551.576 DE-101 550 DE-101 Grundner, Arthur verfasserin (orcid)0000-0002-3765-242X aut Data-driven cloud cover parameterizations for the ICON earth system model using deep learning and symbolic regression Arthur Grundner Bremen 2024 1 Online-Ressource (xi, 140 Seiten) Illustrationen, Diagramme Text txt rdacontent Computermedien c rdamedia Online-Ressource cr rdacarrier Dissertation Universität Bremen 2024 DE-46 Open Access Controlled Vocabulary for Access Rights http://purl.org/coar/access_right/c_abf2 This thesis delves into the improvement of cloud parameterizations in climate models through machine learning trained on coarse-grained output from high-resolution simulations. Utilizing the ICOsahedral Non-hydrostatic (ICON) modeling framework, it specifically targets the enhancement of cloud cover parameterization within the ICON Earth System Model. Three types of neural networks (NNs) differing in vertical locality are developed to estimate cloud cover, with globally trained NNs even applicable to distinct regional simulations. Interpretability analysis exposes model-specific biases and local relationships with the thermodynamic environment. Despite achieving high predictive performance, NNs necessitate post-hoc interpretation tools. To tackle this issue, a combined hierarchical modeling framework incorporating symbolic regression, feature selection, and physical constraints is proposed. The resulting equations, characterized by simplicity and physical consistency, attain performance comparable to NNs while demonstrating superior transferability to other realistic datasets. Our best equation adeptly captures cloud cover distributions across various regimes, notably excelling in representing marine stratocumulus clouds by learning to utilize the vertical relative humidity gradient. This research underscores the potential of deep learning in achieving accurate cloud parameterizations and emphasizes the effective role of symbolic regression in deriving interpretable, consistent equations for cloud cover. DE-46 Namensnennung 4.0 International CC BY 4.0 cc https://creativecommons.org/licenses/by/4.0/ Archivierung/Langzeitarchivierung gewährleistet PEHB XA-DE-HB pdager DE-46 Cloud Cover Parameterization Deep Learning Neural Networks ICON Pareto Frontier PySR Sequential Feature Selection Equation Discovery Physical Constraints Machine Learning Symbolic Regression SHAP Hochschulschrift (DE-588)4113937-9 (DE-627)105825778 (DE-576)209480580 gnd-content Eyring, Veronika akademischer betreuerin (DE-588)1255681446 (DE-627)1799770168 dgs Gentine, Pierre akademischer betreuerin dgs Universität Bremen Grad-verleihende Institution (DE-588)2001386-3 (DE-627)101380429 (DE-576)191575038 dgg Bremen (DE-588)4008135-7 (DE-627)106369636 (DE-576)208874569 uvp Erscheint auch als Druck-Ausgabe Grundner, Arthur Data-driven cloud cover parameterizations for the ICON earth system model using deep learning and symbolic regression Bremen, 2023 xi, 140 Seiten (DE-627)1883908809 https://doi.org/10.26092/elib/2821 Resolving-System kostenfrei https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 Resolving-System kostenfrei https://d-nb.info/1339510391/34 Langzeitarchivierung Nationalbibliothek kostenfrei https://media.suub.uni-bremen.de/handle/elib/7739 Verlag kostenfrei GBV-ODiss GBV_ILN_20 ISIL_DE-84 SYSFLAG_1 GBV_KXP GBV_ILN_21 ISIL_DE-46 GBV_ILN_22 ISIL_DE-18 GBV_ILN_23 ISIL_DE-830 GBV_ILN_30 ISIL_DE-104 GBV_ILN_40 ISIL_DE-7 GBV_ILN_60 ISIL_DE-705 GBV_ILN_63 ISIL_DE-Wim2 GBV_ILN_70 ISIL_DE-89 GBV_ILN_105 ISIL_DE-841 GBV_ILN_110 ISIL_DE-Luen4 GBV_ILN_132 ISIL_DE-959 GBV_ILN_151 ISIL_DE-546 GBV_ILN_161 ISIL_DE-960 GBV_ILN_293 ISIL_DE-960-3 GBV_ILN_370 ISIL_DE-1373 GBV_ILN_2403 ISIL_DE-LFER DSpace BO 20 01 0084 4593082714 x 12-10-24 21 01 0046 4502027553 ebook_2024_dissbremen Kostenloser Zugriff zza 20-03-24 22 01 0018 4593184053 SUBolrd xu 12-10-24 23 01 0830 459323218X olr-d x 12-10-24 30 01 0104 4593278708 z 12-10-24 40 01 0007 4593314887 xsn 12-10-24 60 01 0705 4593373239 OLRD z 12-10-24 63 01 3401 4593426561 ORD x 12-10-24 70 01 0089 4593481716 z 12-10-24 105 01 0841 4593871697 z 12-10-24 110 01 3110 4593582741 x 12-10-24 132 01 0959 4593626374 OLR-DISS x 12-10-24 151 01 0546 4593667224 OLR-ODISS z 12-10-24 161 01 0960 4593691222 ORD z 12-10-24 293 01 3293 4593820650 ORD z 12-10-24 370 01 4370 4593854458 x 12-10-24 2403 01 DE-LFER 4511167990 00 --%%-- --%%-- n --%%-- l01 12-04-24 20 01 0084 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 21 01 0046 https://doi.org/10.26092/elib/2821 LF 22 01 0018 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 23 01 0830 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 30 01 0104 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 40 01 0007 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 60 01 0705 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 63 01 3401 E-Book https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 LF 70 01 0089 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 105 01 0841 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 110 01 3110 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 132 01 0959 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 151 01 0546 Volltext https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 161 01 0960 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 293 01 3293 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 370 01 4370 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 2403 01 DE-LFER https://doi.org/10.26092/elib/2821 21 00 DE-46 00 Universität Bremen 21 00 DE-46 00 Fachbereich 01: Physik/Elektrotechnik (FB 01) 60 01 0705 10 ho 20 01 0084 OLRD 110 01 3110 OLRD 370 01 4370 OLRD 21 01 0046 ebook_2024_dissbremen 22 01 0018 SUBolrd 23 01 0830 olr-d 60 01 0705 OLRD 63 01 3401 ORD 132 01 0959 OLR-DISS 151 01 0546 OLR-ODISS 161 01 0960 ORD 293 01 3293 ORD 23 01 0830 2024-10-12 10:31:16 |
| spelling |
urn:nbn:de:gbv:46-elib77397 urn 10.26092/elib/2821 doi (DE-627)1883908477 (DE-599)KXP1883908477 (OCoLC)1427339102 (OAEPFHB)elib/2821 DE-627 ger DE-627 rda eng XA-DE-HB 551.576 DE-101 550 DE-101 Grundner, Arthur verfasserin (orcid)0000-0002-3765-242X aut Data-driven cloud cover parameterizations for the ICON earth system model using deep learning and symbolic regression Arthur Grundner Bremen 2024 1 Online-Ressource (xi, 140 Seiten) Illustrationen, Diagramme Text txt rdacontent Computermedien c rdamedia Online-Ressource cr rdacarrier Dissertation Universität Bremen 2024 DE-46 Open Access Controlled Vocabulary for Access Rights http://purl.org/coar/access_right/c_abf2 This thesis delves into the improvement of cloud parameterizations in climate models through machine learning trained on coarse-grained output from high-resolution simulations. Utilizing the ICOsahedral Non-hydrostatic (ICON) modeling framework, it specifically targets the enhancement of cloud cover parameterization within the ICON Earth System Model. Three types of neural networks (NNs) differing in vertical locality are developed to estimate cloud cover, with globally trained NNs even applicable to distinct regional simulations. Interpretability analysis exposes model-specific biases and local relationships with the thermodynamic environment. Despite achieving high predictive performance, NNs necessitate post-hoc interpretation tools. To tackle this issue, a combined hierarchical modeling framework incorporating symbolic regression, feature selection, and physical constraints is proposed. The resulting equations, characterized by simplicity and physical consistency, attain performance comparable to NNs while demonstrating superior transferability to other realistic datasets. Our best equation adeptly captures cloud cover distributions across various regimes, notably excelling in representing marine stratocumulus clouds by learning to utilize the vertical relative humidity gradient. This research underscores the potential of deep learning in achieving accurate cloud parameterizations and emphasizes the effective role of symbolic regression in deriving interpretable, consistent equations for cloud cover. DE-46 Namensnennung 4.0 International CC BY 4.0 cc https://creativecommons.org/licenses/by/4.0/ Archivierung/Langzeitarchivierung gewährleistet PEHB XA-DE-HB pdager DE-46 Cloud Cover Parameterization Deep Learning Neural Networks ICON Pareto Frontier PySR Sequential Feature Selection Equation Discovery Physical Constraints Machine Learning Symbolic Regression SHAP Hochschulschrift (DE-588)4113937-9 (DE-627)105825778 (DE-576)209480580 gnd-content Eyring, Veronika akademischer betreuerin (DE-588)1255681446 (DE-627)1799770168 dgs Gentine, Pierre akademischer betreuerin dgs Universität Bremen Grad-verleihende Institution (DE-588)2001386-3 (DE-627)101380429 (DE-576)191575038 dgg Bremen (DE-588)4008135-7 (DE-627)106369636 (DE-576)208874569 uvp Erscheint auch als Druck-Ausgabe Grundner, Arthur Data-driven cloud cover parameterizations for the ICON earth system model using deep learning and symbolic regression Bremen, 2023 xi, 140 Seiten (DE-627)1883908809 https://doi.org/10.26092/elib/2821 Resolving-System kostenfrei https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 Resolving-System kostenfrei https://d-nb.info/1339510391/34 Langzeitarchivierung Nationalbibliothek kostenfrei https://media.suub.uni-bremen.de/handle/elib/7739 Verlag kostenfrei GBV-ODiss GBV_ILN_20 ISIL_DE-84 SYSFLAG_1 GBV_KXP GBV_ILN_21 ISIL_DE-46 GBV_ILN_22 ISIL_DE-18 GBV_ILN_23 ISIL_DE-830 GBV_ILN_30 ISIL_DE-104 GBV_ILN_40 ISIL_DE-7 GBV_ILN_60 ISIL_DE-705 GBV_ILN_63 ISIL_DE-Wim2 GBV_ILN_70 ISIL_DE-89 GBV_ILN_105 ISIL_DE-841 GBV_ILN_110 ISIL_DE-Luen4 GBV_ILN_132 ISIL_DE-959 GBV_ILN_151 ISIL_DE-546 GBV_ILN_161 ISIL_DE-960 GBV_ILN_293 ISIL_DE-960-3 GBV_ILN_370 ISIL_DE-1373 GBV_ILN_2403 ISIL_DE-LFER DSpace BO 20 01 0084 4593082714 x 12-10-24 21 01 0046 4502027553 ebook_2024_dissbremen Kostenloser Zugriff zza 20-03-24 22 01 0018 4593184053 SUBolrd xu 12-10-24 23 01 0830 459323218X olr-d x 12-10-24 30 01 0104 4593278708 z 12-10-24 40 01 0007 4593314887 xsn 12-10-24 60 01 0705 4593373239 OLRD z 12-10-24 63 01 3401 4593426561 ORD x 12-10-24 70 01 0089 4593481716 z 12-10-24 105 01 0841 4593871697 z 12-10-24 110 01 3110 4593582741 x 12-10-24 132 01 0959 4593626374 OLR-DISS x 12-10-24 151 01 0546 4593667224 OLR-ODISS z 12-10-24 161 01 0960 4593691222 ORD z 12-10-24 293 01 3293 4593820650 ORD z 12-10-24 370 01 4370 4593854458 x 12-10-24 2403 01 DE-LFER 4511167990 00 --%%-- --%%-- n --%%-- l01 12-04-24 20 01 0084 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 21 01 0046 https://doi.org/10.26092/elib/2821 LF 22 01 0018 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 23 01 0830 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 30 01 0104 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 40 01 0007 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 60 01 0705 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 63 01 3401 E-Book https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 LF 70 01 0089 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 105 01 0841 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 110 01 3110 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 132 01 0959 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 151 01 0546 Volltext https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 161 01 0960 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 293 01 3293 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 370 01 4370 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 2403 01 DE-LFER https://doi.org/10.26092/elib/2821 21 00 DE-46 00 Universität Bremen 21 00 DE-46 00 Fachbereich 01: Physik/Elektrotechnik (FB 01) 60 01 0705 10 ho 20 01 0084 OLRD 110 01 3110 OLRD 370 01 4370 OLRD 21 01 0046 ebook_2024_dissbremen 22 01 0018 SUBolrd 23 01 0830 olr-d 60 01 0705 OLRD 63 01 3401 ORD 132 01 0959 OLR-DISS 151 01 0546 OLR-ODISS 161 01 0960 ORD 293 01 3293 ORD 23 01 0830 2024-10-12 10:31:16 |
| allfields_unstemmed |
urn:nbn:de:gbv:46-elib77397 urn 10.26092/elib/2821 doi (DE-627)1883908477 (DE-599)KXP1883908477 (OCoLC)1427339102 (OAEPFHB)elib/2821 DE-627 ger DE-627 rda eng XA-DE-HB 551.576 DE-101 550 DE-101 Grundner, Arthur verfasserin (orcid)0000-0002-3765-242X aut Data-driven cloud cover parameterizations for the ICON earth system model using deep learning and symbolic regression Arthur Grundner Bremen 2024 1 Online-Ressource (xi, 140 Seiten) Illustrationen, Diagramme Text txt rdacontent Computermedien c rdamedia Online-Ressource cr rdacarrier Dissertation Universität Bremen 2024 DE-46 Open Access Controlled Vocabulary for Access Rights http://purl.org/coar/access_right/c_abf2 This thesis delves into the improvement of cloud parameterizations in climate models through machine learning trained on coarse-grained output from high-resolution simulations. Utilizing the ICOsahedral Non-hydrostatic (ICON) modeling framework, it specifically targets the enhancement of cloud cover parameterization within the ICON Earth System Model. Three types of neural networks (NNs) differing in vertical locality are developed to estimate cloud cover, with globally trained NNs even applicable to distinct regional simulations. Interpretability analysis exposes model-specific biases and local relationships with the thermodynamic environment. Despite achieving high predictive performance, NNs necessitate post-hoc interpretation tools. To tackle this issue, a combined hierarchical modeling framework incorporating symbolic regression, feature selection, and physical constraints is proposed. The resulting equations, characterized by simplicity and physical consistency, attain performance comparable to NNs while demonstrating superior transferability to other realistic datasets. Our best equation adeptly captures cloud cover distributions across various regimes, notably excelling in representing marine stratocumulus clouds by learning to utilize the vertical relative humidity gradient. This research underscores the potential of deep learning in achieving accurate cloud parameterizations and emphasizes the effective role of symbolic regression in deriving interpretable, consistent equations for cloud cover. DE-46 Namensnennung 4.0 International CC BY 4.0 cc https://creativecommons.org/licenses/by/4.0/ Archivierung/Langzeitarchivierung gewährleistet PEHB XA-DE-HB pdager DE-46 Cloud Cover Parameterization Deep Learning Neural Networks ICON Pareto Frontier PySR Sequential Feature Selection Equation Discovery Physical Constraints Machine Learning Symbolic Regression SHAP Hochschulschrift (DE-588)4113937-9 (DE-627)105825778 (DE-576)209480580 gnd-content Eyring, Veronika akademischer betreuerin (DE-588)1255681446 (DE-627)1799770168 dgs Gentine, Pierre akademischer betreuerin dgs Universität Bremen Grad-verleihende Institution (DE-588)2001386-3 (DE-627)101380429 (DE-576)191575038 dgg Bremen (DE-588)4008135-7 (DE-627)106369636 (DE-576)208874569 uvp Erscheint auch als Druck-Ausgabe Grundner, Arthur Data-driven cloud cover parameterizations for the ICON earth system model using deep learning and symbolic regression Bremen, 2023 xi, 140 Seiten (DE-627)1883908809 https://doi.org/10.26092/elib/2821 Resolving-System kostenfrei https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 Resolving-System kostenfrei https://d-nb.info/1339510391/34 Langzeitarchivierung Nationalbibliothek kostenfrei https://media.suub.uni-bremen.de/handle/elib/7739 Verlag kostenfrei GBV-ODiss GBV_ILN_20 ISIL_DE-84 SYSFLAG_1 GBV_KXP GBV_ILN_21 ISIL_DE-46 GBV_ILN_22 ISIL_DE-18 GBV_ILN_23 ISIL_DE-830 GBV_ILN_30 ISIL_DE-104 GBV_ILN_40 ISIL_DE-7 GBV_ILN_60 ISIL_DE-705 GBV_ILN_63 ISIL_DE-Wim2 GBV_ILN_70 ISIL_DE-89 GBV_ILN_105 ISIL_DE-841 GBV_ILN_110 ISIL_DE-Luen4 GBV_ILN_132 ISIL_DE-959 GBV_ILN_151 ISIL_DE-546 GBV_ILN_161 ISIL_DE-960 GBV_ILN_293 ISIL_DE-960-3 GBV_ILN_370 ISIL_DE-1373 GBV_ILN_2403 ISIL_DE-LFER DSpace BO 20 01 0084 4593082714 x 12-10-24 21 01 0046 4502027553 ebook_2024_dissbremen Kostenloser Zugriff zza 20-03-24 22 01 0018 4593184053 SUBolrd xu 12-10-24 23 01 0830 459323218X olr-d x 12-10-24 30 01 0104 4593278708 z 12-10-24 40 01 0007 4593314887 xsn 12-10-24 60 01 0705 4593373239 OLRD z 12-10-24 63 01 3401 4593426561 ORD x 12-10-24 70 01 0089 4593481716 z 12-10-24 105 01 0841 4593871697 z 12-10-24 110 01 3110 4593582741 x 12-10-24 132 01 0959 4593626374 OLR-DISS x 12-10-24 151 01 0546 4593667224 OLR-ODISS z 12-10-24 161 01 0960 4593691222 ORD z 12-10-24 293 01 3293 4593820650 ORD z 12-10-24 370 01 4370 4593854458 x 12-10-24 2403 01 DE-LFER 4511167990 00 --%%-- --%%-- n --%%-- l01 12-04-24 20 01 0084 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 21 01 0046 https://doi.org/10.26092/elib/2821 LF 22 01 0018 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 23 01 0830 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 30 01 0104 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 40 01 0007 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 60 01 0705 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 63 01 3401 E-Book https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 LF 70 01 0089 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 105 01 0841 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 110 01 3110 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 132 01 0959 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 151 01 0546 Volltext https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 161 01 0960 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 293 01 3293 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 370 01 4370 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 2403 01 DE-LFER https://doi.org/10.26092/elib/2821 21 00 DE-46 00 Universität Bremen 21 00 DE-46 00 Fachbereich 01: Physik/Elektrotechnik (FB 01) 60 01 0705 10 ho 20 01 0084 OLRD 110 01 3110 OLRD 370 01 4370 OLRD 21 01 0046 ebook_2024_dissbremen 22 01 0018 SUBolrd 23 01 0830 olr-d 60 01 0705 OLRD 63 01 3401 ORD 132 01 0959 OLR-DISS 151 01 0546 OLR-ODISS 161 01 0960 ORD 293 01 3293 ORD 23 01 0830 2024-10-12 10:31:16 |
| allfieldsGer |
urn:nbn:de:gbv:46-elib77397 urn 10.26092/elib/2821 doi (DE-627)1883908477 (DE-599)KXP1883908477 (OCoLC)1427339102 (OAEPFHB)elib/2821 DE-627 ger DE-627 rda eng XA-DE-HB 551.576 DE-101 550 DE-101 Grundner, Arthur verfasserin (orcid)0000-0002-3765-242X aut Data-driven cloud cover parameterizations for the ICON earth system model using deep learning and symbolic regression Arthur Grundner Bremen 2024 1 Online-Ressource (xi, 140 Seiten) Illustrationen, Diagramme Text txt rdacontent Computermedien c rdamedia Online-Ressource cr rdacarrier Dissertation Universität Bremen 2024 DE-46 Open Access Controlled Vocabulary for Access Rights http://purl.org/coar/access_right/c_abf2 This thesis delves into the improvement of cloud parameterizations in climate models through machine learning trained on coarse-grained output from high-resolution simulations. Utilizing the ICOsahedral Non-hydrostatic (ICON) modeling framework, it specifically targets the enhancement of cloud cover parameterization within the ICON Earth System Model. Three types of neural networks (NNs) differing in vertical locality are developed to estimate cloud cover, with globally trained NNs even applicable to distinct regional simulations. Interpretability analysis exposes model-specific biases and local relationships with the thermodynamic environment. Despite achieving high predictive performance, NNs necessitate post-hoc interpretation tools. To tackle this issue, a combined hierarchical modeling framework incorporating symbolic regression, feature selection, and physical constraints is proposed. The resulting equations, characterized by simplicity and physical consistency, attain performance comparable to NNs while demonstrating superior transferability to other realistic datasets. Our best equation adeptly captures cloud cover distributions across various regimes, notably excelling in representing marine stratocumulus clouds by learning to utilize the vertical relative humidity gradient. This research underscores the potential of deep learning in achieving accurate cloud parameterizations and emphasizes the effective role of symbolic regression in deriving interpretable, consistent equations for cloud cover. DE-46 Namensnennung 4.0 International CC BY 4.0 cc https://creativecommons.org/licenses/by/4.0/ Archivierung/Langzeitarchivierung gewährleistet PEHB XA-DE-HB pdager DE-46 Cloud Cover Parameterization Deep Learning Neural Networks ICON Pareto Frontier PySR Sequential Feature Selection Equation Discovery Physical Constraints Machine Learning Symbolic Regression SHAP Hochschulschrift (DE-588)4113937-9 (DE-627)105825778 (DE-576)209480580 gnd-content Eyring, Veronika akademischer betreuerin (DE-588)1255681446 (DE-627)1799770168 dgs Gentine, Pierre akademischer betreuerin dgs Universität Bremen Grad-verleihende Institution (DE-588)2001386-3 (DE-627)101380429 (DE-576)191575038 dgg Bremen (DE-588)4008135-7 (DE-627)106369636 (DE-576)208874569 uvp Erscheint auch als Druck-Ausgabe Grundner, Arthur Data-driven cloud cover parameterizations for the ICON earth system model using deep learning and symbolic regression Bremen, 2023 xi, 140 Seiten (DE-627)1883908809 https://doi.org/10.26092/elib/2821 Resolving-System kostenfrei https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 Resolving-System kostenfrei https://d-nb.info/1339510391/34 Langzeitarchivierung Nationalbibliothek kostenfrei https://media.suub.uni-bremen.de/handle/elib/7739 Verlag kostenfrei GBV-ODiss GBV_ILN_20 ISIL_DE-84 SYSFLAG_1 GBV_KXP GBV_ILN_21 ISIL_DE-46 GBV_ILN_22 ISIL_DE-18 GBV_ILN_23 ISIL_DE-830 GBV_ILN_30 ISIL_DE-104 GBV_ILN_40 ISIL_DE-7 GBV_ILN_60 ISIL_DE-705 GBV_ILN_63 ISIL_DE-Wim2 GBV_ILN_70 ISIL_DE-89 GBV_ILN_105 ISIL_DE-841 GBV_ILN_110 ISIL_DE-Luen4 GBV_ILN_132 ISIL_DE-959 GBV_ILN_151 ISIL_DE-546 GBV_ILN_161 ISIL_DE-960 GBV_ILN_293 ISIL_DE-960-3 GBV_ILN_370 ISIL_DE-1373 GBV_ILN_2403 ISIL_DE-LFER DSpace BO 20 01 0084 4593082714 x 12-10-24 21 01 0046 4502027553 ebook_2024_dissbremen Kostenloser Zugriff zza 20-03-24 22 01 0018 4593184053 SUBolrd xu 12-10-24 23 01 0830 459323218X olr-d x 12-10-24 30 01 0104 4593278708 z 12-10-24 40 01 0007 4593314887 xsn 12-10-24 60 01 0705 4593373239 OLRD z 12-10-24 63 01 3401 4593426561 ORD x 12-10-24 70 01 0089 4593481716 z 12-10-24 105 01 0841 4593871697 z 12-10-24 110 01 3110 4593582741 x 12-10-24 132 01 0959 4593626374 OLR-DISS x 12-10-24 151 01 0546 4593667224 OLR-ODISS z 12-10-24 161 01 0960 4593691222 ORD z 12-10-24 293 01 3293 4593820650 ORD z 12-10-24 370 01 4370 4593854458 x 12-10-24 2403 01 DE-LFER 4511167990 00 --%%-- --%%-- n --%%-- l01 12-04-24 20 01 0084 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 21 01 0046 https://doi.org/10.26092/elib/2821 LF 22 01 0018 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 23 01 0830 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 30 01 0104 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 40 01 0007 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 60 01 0705 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 63 01 3401 E-Book https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 LF 70 01 0089 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 105 01 0841 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 110 01 3110 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 132 01 0959 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 151 01 0546 Volltext https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 161 01 0960 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 293 01 3293 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 370 01 4370 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 2403 01 DE-LFER https://doi.org/10.26092/elib/2821 21 00 DE-46 00 Universität Bremen 21 00 DE-46 00 Fachbereich 01: Physik/Elektrotechnik (FB 01) 60 01 0705 10 ho 20 01 0084 OLRD 110 01 3110 OLRD 370 01 4370 OLRD 21 01 0046 ebook_2024_dissbremen 22 01 0018 SUBolrd 23 01 0830 olr-d 60 01 0705 OLRD 63 01 3401 ORD 132 01 0959 OLR-DISS 151 01 0546 OLR-ODISS 161 01 0960 ORD 293 01 3293 ORD 23 01 0830 2024-10-12 10:31:16 |
| allfieldsSound |
urn:nbn:de:gbv:46-elib77397 urn 10.26092/elib/2821 doi (DE-627)1883908477 (DE-599)KXP1883908477 (OCoLC)1427339102 (OAEPFHB)elib/2821 DE-627 ger DE-627 rda eng XA-DE-HB 551.576 DE-101 550 DE-101 Grundner, Arthur verfasserin (orcid)0000-0002-3765-242X aut Data-driven cloud cover parameterizations for the ICON earth system model using deep learning and symbolic regression Arthur Grundner Bremen 2024 1 Online-Ressource (xi, 140 Seiten) Illustrationen, Diagramme Text txt rdacontent Computermedien c rdamedia Online-Ressource cr rdacarrier Dissertation Universität Bremen 2024 DE-46 Open Access Controlled Vocabulary for Access Rights http://purl.org/coar/access_right/c_abf2 This thesis delves into the improvement of cloud parameterizations in climate models through machine learning trained on coarse-grained output from high-resolution simulations. Utilizing the ICOsahedral Non-hydrostatic (ICON) modeling framework, it specifically targets the enhancement of cloud cover parameterization within the ICON Earth System Model. Three types of neural networks (NNs) differing in vertical locality are developed to estimate cloud cover, with globally trained NNs even applicable to distinct regional simulations. Interpretability analysis exposes model-specific biases and local relationships with the thermodynamic environment. Despite achieving high predictive performance, NNs necessitate post-hoc interpretation tools. To tackle this issue, a combined hierarchical modeling framework incorporating symbolic regression, feature selection, and physical constraints is proposed. The resulting equations, characterized by simplicity and physical consistency, attain performance comparable to NNs while demonstrating superior transferability to other realistic datasets. Our best equation adeptly captures cloud cover distributions across various regimes, notably excelling in representing marine stratocumulus clouds by learning to utilize the vertical relative humidity gradient. This research underscores the potential of deep learning in achieving accurate cloud parameterizations and emphasizes the effective role of symbolic regression in deriving interpretable, consistent equations for cloud cover. DE-46 Namensnennung 4.0 International CC BY 4.0 cc https://creativecommons.org/licenses/by/4.0/ Archivierung/Langzeitarchivierung gewährleistet PEHB XA-DE-HB pdager DE-46 Cloud Cover Parameterization Deep Learning Neural Networks ICON Pareto Frontier PySR Sequential Feature Selection Equation Discovery Physical Constraints Machine Learning Symbolic Regression SHAP Hochschulschrift (DE-588)4113937-9 (DE-627)105825778 (DE-576)209480580 gnd-content Eyring, Veronika akademischer betreuerin (DE-588)1255681446 (DE-627)1799770168 dgs Gentine, Pierre akademischer betreuerin dgs Universität Bremen Grad-verleihende Institution (DE-588)2001386-3 (DE-627)101380429 (DE-576)191575038 dgg Bremen (DE-588)4008135-7 (DE-627)106369636 (DE-576)208874569 uvp Erscheint auch als Druck-Ausgabe Grundner, Arthur Data-driven cloud cover parameterizations for the ICON earth system model using deep learning and symbolic regression Bremen, 2023 xi, 140 Seiten (DE-627)1883908809 https://doi.org/10.26092/elib/2821 Resolving-System kostenfrei https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 Resolving-System kostenfrei https://d-nb.info/1339510391/34 Langzeitarchivierung Nationalbibliothek kostenfrei https://media.suub.uni-bremen.de/handle/elib/7739 Verlag kostenfrei GBV-ODiss GBV_ILN_20 ISIL_DE-84 SYSFLAG_1 GBV_KXP GBV_ILN_21 ISIL_DE-46 GBV_ILN_22 ISIL_DE-18 GBV_ILN_23 ISIL_DE-830 GBV_ILN_30 ISIL_DE-104 GBV_ILN_40 ISIL_DE-7 GBV_ILN_60 ISIL_DE-705 GBV_ILN_63 ISIL_DE-Wim2 GBV_ILN_70 ISIL_DE-89 GBV_ILN_105 ISIL_DE-841 GBV_ILN_110 ISIL_DE-Luen4 GBV_ILN_132 ISIL_DE-959 GBV_ILN_151 ISIL_DE-546 GBV_ILN_161 ISIL_DE-960 GBV_ILN_293 ISIL_DE-960-3 GBV_ILN_370 ISIL_DE-1373 GBV_ILN_2403 ISIL_DE-LFER DSpace BO 20 01 0084 4593082714 x 12-10-24 21 01 0046 4502027553 ebook_2024_dissbremen Kostenloser Zugriff zza 20-03-24 22 01 0018 4593184053 SUBolrd xu 12-10-24 23 01 0830 459323218X olr-d x 12-10-24 30 01 0104 4593278708 z 12-10-24 40 01 0007 4593314887 xsn 12-10-24 60 01 0705 4593373239 OLRD z 12-10-24 63 01 3401 4593426561 ORD x 12-10-24 70 01 0089 4593481716 z 12-10-24 105 01 0841 4593871697 z 12-10-24 110 01 3110 4593582741 x 12-10-24 132 01 0959 4593626374 OLR-DISS x 12-10-24 151 01 0546 4593667224 OLR-ODISS z 12-10-24 161 01 0960 4593691222 ORD z 12-10-24 293 01 3293 4593820650 ORD z 12-10-24 370 01 4370 4593854458 x 12-10-24 2403 01 DE-LFER 4511167990 00 --%%-- --%%-- n --%%-- l01 12-04-24 20 01 0084 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 21 01 0046 https://doi.org/10.26092/elib/2821 LF 22 01 0018 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 23 01 0830 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 30 01 0104 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 40 01 0007 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 60 01 0705 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 63 01 3401 E-Book https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 LF 70 01 0089 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 105 01 0841 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 110 01 3110 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 132 01 0959 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 151 01 0546 Volltext https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 161 01 0960 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 293 01 3293 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 370 01 4370 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 2403 01 DE-LFER https://doi.org/10.26092/elib/2821 21 00 DE-46 00 Universität Bremen 21 00 DE-46 00 Fachbereich 01: Physik/Elektrotechnik (FB 01) 60 01 0705 10 ho 20 01 0084 OLRD 110 01 3110 OLRD 370 01 4370 OLRD 21 01 0046 ebook_2024_dissbremen 22 01 0018 SUBolrd 23 01 0830 olr-d 60 01 0705 OLRD 63 01 3401 ORD 132 01 0959 OLR-DISS 151 01 0546 OLR-ODISS 161 01 0960 ORD 293 01 3293 ORD 23 01 0830 2024-10-12 10:31:16 |
| language |
English |
| format_phy_str_mv |
Book |
| building |
20 21:z 22:u 23 30 40:s 60 63 70 105 110 132 151 161 293 370 2403:0 |
| institution |
findex.gbv.de |
| selectbib_iln_str_mv |
20@ 21@za 22@u 23@ 30@ 40@sn 60@ 63@ 70@ 105@ 110@ 132@ 151@ 161@ 293@ 370@ 2403@01 |
| topic_facet |
Cloud Cover Parameterization Deep Learning Neural Networks ICON Pareto Frontier PySR Sequential Feature Selection Equation Discovery Physical Constraints Machine Learning Symbolic Regression SHAP |
| hochschulschrift_txt_mv |
Dissertation Universität Bremen 2024 |
| sw_local_iln_str_mv |
21:Universität Bremen DE-46:Universität Bremen 21:Fachbereich 01: Physik/Elektrotechnik (FB 01) DE-46:Fachbereich 01: Physik/Elektrotechnik (FB 01) |
| dewey-raw |
551.576 |
| isfreeaccess_bool |
true |
| authorswithroles_txt_mv |
Grundner, Arthur @@aut@@ Eyring, Veronika @@dgs@@ Gentine, Pierre @@dgs@@ |
| publishDateDaySort_date |
2024-01-01T00:00:00Z |
| dewey-sort |
3551.576 |
| id |
1883908477 |
| genre_facet |
Hochschulschrift |
| language_de |
englisch |
| fullrecord |
<?xml version="1.0" encoding="UTF-8"?><collection xmlns="http://www.loc.gov/MARC21/slim"><record><leader>01000cam a2200265 4500</leader><controlfield tag="001">1883908477</controlfield><controlfield tag="003">DE-627</controlfield><controlfield tag="005">20240912032145.0</controlfield><controlfield tag="007">cr uuu---uuuuu</controlfield><controlfield tag="008">240320s2024 gw |||||om 00| ||eng c</controlfield><datafield tag="024" ind1="7" ind2=" "><subfield code="a">urn:nbn:de:gbv:46-elib77397</subfield><subfield code="2">urn</subfield></datafield><datafield tag="024" ind1="7" ind2=" "><subfield code="a">10.26092/elib/2821</subfield><subfield code="2">doi</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-627)1883908477</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-599)KXP1883908477</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(OCoLC)1427339102</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(OAEPFHB)elib/2821</subfield></datafield><datafield tag="040" ind1=" " ind2=" "><subfield code="a">DE-627</subfield><subfield code="b">ger</subfield><subfield code="c">DE-627</subfield><subfield code="e">rda</subfield></datafield><datafield tag="041" ind1=" " ind2=" "><subfield code="a">eng</subfield></datafield><datafield tag="044" ind1=" " ind2=" "><subfield code="c">XA-DE-HB</subfield></datafield><datafield tag="082" ind1="0" ind2="4"><subfield code="a">551.576</subfield><subfield code="q">DE-101</subfield></datafield><datafield tag="082" ind1="0" ind2="4"><subfield code="a">550</subfield><subfield code="q">DE-101</subfield></datafield><datafield tag="100" ind1="1" ind2=" "><subfield code="a">Grundner, Arthur</subfield><subfield code="e">verfasserin</subfield><subfield code="0">(orcid)0000-0002-3765-242X</subfield><subfield code="4">aut</subfield></datafield><datafield tag="245" ind1="1" ind2="0"><subfield code="a">Data-driven cloud cover parameterizations for the ICON earth system model using deep learning and symbolic regression</subfield><subfield code="c">Arthur Grundner</subfield></datafield><datafield tag="264" ind1=" " ind2="1"><subfield code="a">Bremen</subfield><subfield code="c">2024</subfield></datafield><datafield tag="300" ind1=" " ind2=" "><subfield code="a">1 Online-Ressource (xi, 140 Seiten)</subfield><subfield code="b">Illustrationen, Diagramme</subfield></datafield><datafield tag="336" ind1=" " ind2=" "><subfield code="a">Text</subfield><subfield code="b">txt</subfield><subfield code="2">rdacontent</subfield></datafield><datafield tag="337" ind1=" " ind2=" "><subfield code="a">Computermedien</subfield><subfield code="b">c</subfield><subfield code="2">rdamedia</subfield></datafield><datafield tag="338" ind1=" " ind2=" "><subfield code="a">Online-Ressource</subfield><subfield code="b">cr</subfield><subfield code="2">rdacarrier</subfield></datafield><datafield tag="502" ind1=" " ind2=" "><subfield code="b">Dissertation</subfield><subfield code="c">Universität Bremen</subfield><subfield code="d">2024</subfield></datafield><datafield tag="506" ind1="0" ind2=" "><subfield code="q">DE-46</subfield><subfield code="a">Open Access</subfield><subfield code="e">Controlled Vocabulary for Access Rights</subfield><subfield code="u">http://purl.org/coar/access_right/c_abf2</subfield></datafield><datafield tag="520" ind1=" " ind2=" "><subfield code="a">This thesis delves into the improvement of cloud parameterizations in climate models through machine learning trained on coarse-grained output from high-resolution simulations. Utilizing the ICOsahedral Non-hydrostatic (ICON) modeling framework, it specifically targets the enhancement of cloud cover parameterization within the ICON Earth System Model. Three types of neural networks (NNs) differing in vertical locality are developed to estimate cloud cover, with globally trained NNs even applicable to distinct regional simulations. Interpretability analysis exposes model-specific biases and local relationships with the thermodynamic environment. Despite achieving high predictive performance, NNs necessitate post-hoc interpretation tools. To tackle this issue, a combined hierarchical modeling framework incorporating symbolic regression, feature selection, and physical constraints is proposed. The resulting equations, characterized by simplicity and physical consistency, attain performance comparable to NNs while demonstrating superior transferability to other realistic datasets. Our best equation adeptly captures cloud cover distributions across various regimes, notably excelling in representing marine stratocumulus clouds by learning to utilize the vertical relative humidity gradient. This research underscores the potential of deep learning in achieving accurate cloud parameterizations and emphasizes the effective role of symbolic regression in deriving interpretable, consistent equations for cloud cover.</subfield></datafield><datafield tag="540" ind1=" " ind2=" "><subfield code="q">DE-46</subfield><subfield code="a">Namensnennung 4.0 International</subfield><subfield code="f">CC BY 4.0</subfield><subfield code="2">cc</subfield><subfield code="u">https://creativecommons.org/licenses/by/4.0/</subfield></datafield><datafield tag="583" ind1="1" ind2=" "><subfield code="a">Archivierung/Langzeitarchivierung gewährleistet</subfield><subfield code="f">PEHB</subfield><subfield code="x">XA-DE-HB</subfield><subfield code="2">pdager</subfield><subfield code="5">DE-46</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Cloud Cover</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Parameterization</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Deep Learning</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Neural Networks</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">ICON</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Pareto Frontier</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">PySR</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Sequential Feature Selection</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Equation Discovery</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Physical Constraints</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Machine Learning</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Symbolic Regression</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">SHAP</subfield></datafield><datafield tag="655" ind1=" " ind2="7"><subfield code="a">Hochschulschrift</subfield><subfield code="0">(DE-588)4113937-9</subfield><subfield code="0">(DE-627)105825778</subfield><subfield code="0">(DE-576)209480580</subfield><subfield code="2">gnd-content</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Eyring, Veronika</subfield><subfield code="e">akademischer betreuerin</subfield><subfield code="0">(DE-588)1255681446</subfield><subfield code="0">(DE-627)1799770168</subfield><subfield code="4">dgs</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Gentine, Pierre</subfield><subfield code="e">akademischer betreuerin</subfield><subfield code="4">dgs</subfield></datafield><datafield tag="710" ind1="2" ind2=" "><subfield code="a">Universität Bremen</subfield><subfield code="e">Grad-verleihende Institution</subfield><subfield code="0">(DE-588)2001386-3</subfield><subfield code="0">(DE-627)101380429</subfield><subfield code="0">(DE-576)191575038</subfield><subfield code="4">dgg</subfield></datafield><datafield tag="751" ind1=" " ind2=" "><subfield code="a">Bremen</subfield><subfield code="0">(DE-588)4008135-7</subfield><subfield code="0">(DE-627)106369636</subfield><subfield code="0">(DE-576)208874569</subfield><subfield code="4">uvp</subfield></datafield><datafield tag="776" ind1="0" ind2="8"><subfield code="i">Erscheint auch als</subfield><subfield code="n">Druck-Ausgabe</subfield><subfield code="a">Grundner, Arthur</subfield><subfield code="t">Data-driven cloud cover parameterizations for the ICON earth system model using deep learning and symbolic regression</subfield><subfield code="d">Bremen, 2023</subfield><subfield code="h">xi, 140 Seiten</subfield><subfield code="w">(DE-627)1883908809</subfield></datafield><datafield tag="856" ind1="4" ind2="0"><subfield code="u">https://doi.org/10.26092/elib/2821</subfield><subfield code="x">Resolving-System</subfield><subfield code="z">kostenfrei</subfield></datafield><datafield tag="856" ind1="4" ind2="0"><subfield code="u">https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397</subfield><subfield code="x">Resolving-System</subfield><subfield code="z">kostenfrei</subfield></datafield><datafield tag="856" ind1="4" ind2="0"><subfield code="u">https://d-nb.info/1339510391/34</subfield><subfield code="x">Langzeitarchivierung Nationalbibliothek</subfield><subfield code="z">kostenfrei</subfield></datafield><datafield tag="856" ind1="4" ind2="0"><subfield code="u">https://media.suub.uni-bremen.de/handle/elib/7739</subfield><subfield code="x">Verlag</subfield><subfield code="z">kostenfrei</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV-ODiss</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_20</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-84</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SYSFLAG_1</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_KXP</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_21</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-46</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_22</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-18</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_23</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-830</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_30</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-104</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_40</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-7</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_60</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-705</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_63</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-Wim2</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_70</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-89</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_105</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-841</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_110</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-Luen4</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_132</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-959</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_151</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-546</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_161</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-960</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_293</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-960-3</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_370</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-1373</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2403</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-LFER</subfield></datafield><datafield tag="935" ind1=" " ind2=" "><subfield code="i">DSpace</subfield></datafield><datafield tag="951" ind1=" " ind2=" "><subfield code="a">BO</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">20</subfield><subfield code="1">01</subfield><subfield code="x">0084</subfield><subfield code="b">4593082714</subfield><subfield code="y">x</subfield><subfield code="z">12-10-24</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">21</subfield><subfield code="1">01</subfield><subfield code="x">0046</subfield><subfield code="b">4502027553</subfield><subfield code="h">ebook_2024_dissbremen</subfield><subfield code="k">Kostenloser Zugriff</subfield><subfield code="y">zza</subfield><subfield code="z">20-03-24</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">22</subfield><subfield code="1">01</subfield><subfield code="x">0018</subfield><subfield code="b">4593184053</subfield><subfield code="h">SUBolrd</subfield><subfield code="y">xu</subfield><subfield code="z">12-10-24</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">23</subfield><subfield code="1">01</subfield><subfield code="x">0830</subfield><subfield code="b">459323218X</subfield><subfield code="h">olr-d</subfield><subfield code="y">x</subfield><subfield code="z">12-10-24</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">30</subfield><subfield code="1">01</subfield><subfield code="x">0104</subfield><subfield code="b">4593278708</subfield><subfield code="y">z</subfield><subfield code="z">12-10-24</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">40</subfield><subfield code="1">01</subfield><subfield code="x">0007</subfield><subfield code="b">4593314887</subfield><subfield code="y">xsn</subfield><subfield code="z">12-10-24</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">60</subfield><subfield code="1">01</subfield><subfield code="x">0705</subfield><subfield code="b">4593373239</subfield><subfield code="h">OLRD</subfield><subfield code="y">z</subfield><subfield code="z">12-10-24</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">63</subfield><subfield code="1">01</subfield><subfield code="x">3401</subfield><subfield code="b">4593426561</subfield><subfield code="h">ORD</subfield><subfield code="y">x</subfield><subfield code="z">12-10-24</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">70</subfield><subfield code="1">01</subfield><subfield code="x">0089</subfield><subfield code="b">4593481716</subfield><subfield code="y">z</subfield><subfield code="z">12-10-24</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">105</subfield><subfield code="1">01</subfield><subfield code="x">0841</subfield><subfield code="b">4593871697</subfield><subfield code="y">z</subfield><subfield code="z">12-10-24</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">110</subfield><subfield code="1">01</subfield><subfield code="x">3110</subfield><subfield code="b">4593582741</subfield><subfield code="y">x</subfield><subfield code="z">12-10-24</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">132</subfield><subfield code="1">01</subfield><subfield code="x">0959</subfield><subfield code="b">4593626374</subfield><subfield code="h">OLR-DISS</subfield><subfield code="y">x</subfield><subfield code="z">12-10-24</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">151</subfield><subfield code="1">01</subfield><subfield code="x">0546</subfield><subfield code="b">4593667224</subfield><subfield code="h">OLR-ODISS</subfield><subfield code="y">z</subfield><subfield code="z">12-10-24</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">161</subfield><subfield code="1">01</subfield><subfield code="x">0960</subfield><subfield code="b">4593691222</subfield><subfield code="h">ORD</subfield><subfield code="y">z</subfield><subfield code="z">12-10-24</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">293</subfield><subfield code="1">01</subfield><subfield code="x">3293</subfield><subfield code="b">4593820650</subfield><subfield code="h">ORD</subfield><subfield code="y">z</subfield><subfield code="z">12-10-24</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">370</subfield><subfield code="1">01</subfield><subfield code="x">4370</subfield><subfield code="b">4593854458</subfield><subfield code="y">x</subfield><subfield code="z">12-10-24</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">2403</subfield><subfield code="1">01</subfield><subfield code="x">DE-LFER</subfield><subfield code="b">4511167990</subfield><subfield code="c">00</subfield><subfield code="f">--%%--</subfield><subfield code="d">--%%--</subfield><subfield code="e">n</subfield><subfield code="j">--%%--</subfield><subfield code="y">l01</subfield><subfield code="z">12-04-24</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">20</subfield><subfield code="1">01</subfield><subfield code="x">0084</subfield><subfield code="r">https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">21</subfield><subfield code="1">01</subfield><subfield code="x">0046</subfield><subfield code="r">https://doi.org/10.26092/elib/2821</subfield><subfield code="z">LF</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">22</subfield><subfield code="1">01</subfield><subfield code="x">0018</subfield><subfield code="r">https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">23</subfield><subfield code="1">01</subfield><subfield code="x">0830</subfield><subfield code="r">https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">30</subfield><subfield code="1">01</subfield><subfield code="x">0104</subfield><subfield code="r">https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">40</subfield><subfield code="1">01</subfield><subfield code="x">0007</subfield><subfield code="r">https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">60</subfield><subfield code="1">01</subfield><subfield code="x">0705</subfield><subfield code="r">https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">63</subfield><subfield code="1">01</subfield><subfield code="x">3401</subfield><subfield code="y">E-Book</subfield><subfield code="r">https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397</subfield><subfield code="z">LF</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">70</subfield><subfield code="1">01</subfield><subfield code="x">0089</subfield><subfield code="r">https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">105</subfield><subfield code="1">01</subfield><subfield code="x">0841</subfield><subfield code="r">https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">110</subfield><subfield code="1">01</subfield><subfield code="x">3110</subfield><subfield code="r">https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">132</subfield><subfield code="1">01</subfield><subfield code="x">0959</subfield><subfield code="r">https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">151</subfield><subfield code="1">01</subfield><subfield code="x">0546</subfield><subfield code="y">Volltext</subfield><subfield code="r">https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">161</subfield><subfield code="1">01</subfield><subfield code="x">0960</subfield><subfield code="r">https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">293</subfield><subfield code="1">01</subfield><subfield code="x">3293</subfield><subfield code="r">https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">370</subfield><subfield code="1">01</subfield><subfield code="x">4370</subfield><subfield code="r">https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">2403</subfield><subfield code="1">01</subfield><subfield code="x">DE-LFER</subfield><subfield code="r">https://doi.org/10.26092/elib/2821</subfield></datafield><datafield tag="982" ind1=" " ind2=" "><subfield code="2">21</subfield><subfield code="1">00</subfield><subfield code="x">DE-46</subfield><subfield code="8">00</subfield><subfield code="a">Universität Bremen</subfield></datafield><datafield tag="982" ind1=" " ind2=" "><subfield code="2">21</subfield><subfield code="1">00</subfield><subfield code="x">DE-46</subfield><subfield code="8">00</subfield><subfield code="a">Fachbereich 01: Physik/Elektrotechnik (FB 01)</subfield></datafield><datafield tag="983" ind1=" " ind2=" "><subfield code="2">60</subfield><subfield code="1">01</subfield><subfield code="x">0705</subfield><subfield code="8">10</subfield><subfield code="a">ho</subfield></datafield><datafield tag="985" ind1=" " ind2=" "><subfield code="2">20</subfield><subfield code="1">01</subfield><subfield code="x">0084</subfield><subfield code="a">OLRD</subfield></datafield><datafield tag="985" ind1=" " ind2=" "><subfield code="2">110</subfield><subfield code="1">01</subfield><subfield code="x">3110</subfield><subfield code="a">OLRD</subfield></datafield><datafield tag="985" ind1=" " ind2=" "><subfield code="2">370</subfield><subfield code="1">01</subfield><subfield code="x">4370</subfield><subfield code="a">OLRD</subfield></datafield><datafield tag="995" ind1=" " ind2=" "><subfield code="2">21</subfield><subfield code="1">01</subfield><subfield code="x">0046</subfield><subfield code="a">ebook_2024_dissbremen</subfield></datafield><datafield tag="995" ind1=" " ind2=" "><subfield code="2">22</subfield><subfield code="1">01</subfield><subfield code="x">0018</subfield><subfield code="a">SUBolrd</subfield></datafield><datafield tag="995" ind1=" " ind2=" "><subfield code="2">23</subfield><subfield code="1">01</subfield><subfield code="x">0830</subfield><subfield code="a">olr-d</subfield></datafield><datafield tag="995" ind1=" " ind2=" "><subfield code="2">60</subfield><subfield code="1">01</subfield><subfield code="x">0705</subfield><subfield code="a">OLRD</subfield></datafield><datafield tag="995" ind1=" " ind2=" "><subfield code="2">63</subfield><subfield code="1">01</subfield><subfield code="x">3401</subfield><subfield code="a">ORD</subfield></datafield><datafield tag="995" ind1=" " ind2=" "><subfield code="2">132</subfield><subfield code="1">01</subfield><subfield code="x">0959</subfield><subfield code="a">OLR-DISS</subfield></datafield><datafield tag="995" ind1=" " ind2=" "><subfield code="2">151</subfield><subfield code="1">01</subfield><subfield code="x">0546</subfield><subfield code="a">OLR-ODISS</subfield></datafield><datafield tag="995" ind1=" " ind2=" "><subfield code="2">161</subfield><subfield code="1">01</subfield><subfield code="x">0960</subfield><subfield code="a">ORD</subfield></datafield><datafield tag="995" ind1=" " ind2=" "><subfield code="2">293</subfield><subfield code="1">01</subfield><subfield code="x">3293</subfield><subfield code="a">ORD</subfield></datafield><datafield tag="998" ind1=" " ind2=" "><subfield code="2">23</subfield><subfield code="1">01</subfield><subfield code="x">0830</subfield><subfield code="0">2024-10-12 10:31:16</subfield></datafield></record></collection>
|
| standort_str_mv |
--%%-- |
| standort_iln_str_mv |
2403:--%%-- DE-LFER:--%%-- |
| author |
Grundner, Arthur |
| spellingShingle |
Grundner, Arthur ddc 551.576 ddc 550 misc Cloud Cover misc Parameterization misc Deep Learning misc Neural Networks misc ICON misc Pareto Frontier misc PySR misc Sequential Feature Selection misc Equation Discovery misc Physical Constraints misc Machine Learning misc Symbolic Regression misc SHAP 21 Universität Bremen 21 Fachbereich 01: Physik/Elektrotechnik (FB 01) Data-driven cloud cover parameterizations for the ICON earth system model using deep learning and symbolic regression |
| authorStr |
Grundner, Arthur |
| ppnlink_with_tag_str_mv |
@@776@@(DE-627)1883908809 |
| format |
eBook |
| dewey-ones |
551 - Geology, hydrology & meteorology 550 - Earth sciences |
| delete_txt_mv |
keep |
| author_role |
aut |
| collection |
KXP GVK SWB |
| publishPlace |
Bremen |
| remote_str |
true |
| abrufzeichen_iln_str_mv |
21@ebook_2024_dissbremen 22@SUBolrd 23@olr-d 60@OLRD 63@ORD 132@OLR-DISS 151@OLR-ODISS 161@ORD 293@ORD |
| abrufzeichen_iln_scis_mv |
21@ebook_2024_dissbremen 22@SUBolrd 23@olr-d 60@OLRD 63@ORD 132@OLR-DISS 151@OLR-ODISS 161@ORD 293@ORD |
| last_changed_iln_str_mv |
20@12-10-24 21@20-03-24 22@12-10-24 23@12-10-24 30@12-10-24 40@12-10-24 60@12-10-24 63@12-10-24 70@12-10-24 105@12-10-24 110@12-10-24 132@12-10-24 151@12-10-24 161@12-10-24 293@12-10-24 370@12-10-24 2403@12-04-24 |
| illustrated |
Not Illustrated |
| notation_local_iln_str_mv |
60:ho 0705:ho |
| topic_title |
551.576 DE-101 550 DE-101 21 00 DE-46 00 Universität Bremen 21 00 DE-46 00 Fachbereich 01: Physik/Elektrotechnik (FB 01) Data-driven cloud cover parameterizations for the ICON earth system model using deep learning and symbolic regression Arthur Grundner Cloud Cover Parameterization Deep Learning Neural Networks ICON Pareto Frontier PySR Sequential Feature Selection Equation Discovery Physical Constraints Machine Learning Symbolic Regression SHAP |
| topic |
ddc 551.576 ddc 550 misc Cloud Cover misc Parameterization misc Deep Learning misc Neural Networks misc ICON misc Pareto Frontier misc PySR misc Sequential Feature Selection misc Equation Discovery misc Physical Constraints misc Machine Learning misc Symbolic Regression misc SHAP 21 Universität Bremen 21 Fachbereich 01: Physik/Elektrotechnik (FB 01) |
| topic_unstemmed |
ddc 551.576 ddc 550 misc Cloud Cover misc Parameterization misc Deep Learning misc Neural Networks misc ICON misc Pareto Frontier misc PySR misc Sequential Feature Selection misc Equation Discovery misc Physical Constraints misc Machine Learning misc Symbolic Regression misc SHAP 21 Universität Bremen 21 Fachbereich 01: Physik/Elektrotechnik (FB 01) |
| topic_browse |
ddc 551.576 ddc 550 misc Cloud Cover misc Parameterization misc Deep Learning misc Neural Networks misc ICON misc Pareto Frontier misc PySR misc Sequential Feature Selection misc Equation Discovery misc Physical Constraints misc Machine Learning misc Symbolic Regression misc SHAP 21 Universität Bremen 21 Fachbereich 01: Physik/Elektrotechnik (FB 01) |
| format_facet |
Elektronische Bücher Bücher Elektronische Ressource Hochschulschriften |
| standort_txtP_mv |
--%%-- |
| format_main_str_mv |
Text Buch |
| format_details_str_mv |
Hochschulschrift |
| carriertype_str_mv |
cr |
| author2_variant |
v e ve p g pg |
| signature |
--%%-- |
| signature_str_mv |
--%%-- |
| dewey-tens |
550 - Earth sciences & geology |
| isfreeaccess_txt |
true |
| normlinkwithrole_str_mv |
(DE-588)4113937-9@@655@@ (DE-588)1255681446@@dgs@@ (DE-588)2001386-3@@dgg@@ (DE-588)4008135-7@@uvp@@ |
| title |
Data-driven cloud cover parameterizations for the ICON earth system model using deep learning and symbolic regression |
| ctrlnum |
(DE-627)1883908477 (DE-599)KXP1883908477 (OCoLC)1427339102 (OAEPFHB)elib/2821 |
| exemplarkommentar_str_mv |
21@Kostenloser Zugriff |
| title_full |
Data-driven cloud cover parameterizations for the ICON earth system model using deep learning and symbolic regression Arthur Grundner |
| author_sort |
Grundner, Arthur |
| callnumber-first-code |
- |
| lang_code |
eng |
| class_local |
60 01 0705 10 ho |
| selektneu_str_mv |
23@2024-10-12 10:31:16 |
| isOA_bool |
true |
| dewey-hundreds |
500 - Science |
| recordtype |
marc |
| genre |
Hochschulschrift (DE-588)4113937-9 (DE-627)105825778 (DE-576)209480580 gnd-content |
| publishDateSort |
2024 |
| contenttype_str_mv |
txt |
| class_local_iln |
60:ho 0705:ho |
| author_browse |
Grundner, Arthur |
| selectkey |
20:x 21:z 22:x 23:x 30:z 40:x 60:z 63:x 70:z 105:z 110:x 132:x 151:z 161:z 293:z 370:x 2403:l |
| physical |
1 Online-Ressource (xi, 140 Seiten) Illustrationen, Diagramme |
| class |
551.576 DE-101 550 DE-101 60 01 0705 10 ho |
| format_se |
Elektronische Bücher |
| countryofpublication_str_mv |
XA-DE-HB |
| author-letter |
Grundner, Arthur |
| classname_local_iln_str_mv |
60: 0705: |
| doi_str_mv |
10.26092/elib/2821 |
| normlink |
(ORCID)0000-0002-3765-242X 4113937-9 105825778 209480580 1255681446 1799770168 2001386-3 101380429 191575038 4008135-7 106369636 208874569 2024-10-12 10:31:16 |
| normlink_prefix_str_mv |
(orcid)0000-0002-3765-242X (DE-588)4113937-9 (DE-627)105825778 (DE-576)209480580 (DE-588)1255681446 (DE-627)1799770168 (DE-588)2001386-3 (DE-627)101380429 (DE-576)191575038 (DE-588)4008135-7 (DE-627)106369636 (DE-576)208874569 2024-10-12 10:31:16 |
| dewey-full |
551.576 550 |
| author2-role |
akademischer betreuerin Grad-verleihende Institution |
| title_sort |
data-driven cloud cover parameterizations for the icon earth system model using deep learning and symbolic regression |
| title_auth |
Data-driven cloud cover parameterizations for the ICON earth system model using deep learning and symbolic regression |
| abstract |
This thesis delves into the improvement of cloud parameterizations in climate models through machine learning trained on coarse-grained output from high-resolution simulations. Utilizing the ICOsahedral Non-hydrostatic (ICON) modeling framework, it specifically targets the enhancement of cloud cover parameterization within the ICON Earth System Model. Three types of neural networks (NNs) differing in vertical locality are developed to estimate cloud cover, with globally trained NNs even applicable to distinct regional simulations. Interpretability analysis exposes model-specific biases and local relationships with the thermodynamic environment. Despite achieving high predictive performance, NNs necessitate post-hoc interpretation tools. To tackle this issue, a combined hierarchical modeling framework incorporating symbolic regression, feature selection, and physical constraints is proposed. The resulting equations, characterized by simplicity and physical consistency, attain performance comparable to NNs while demonstrating superior transferability to other realistic datasets. Our best equation adeptly captures cloud cover distributions across various regimes, notably excelling in representing marine stratocumulus clouds by learning to utilize the vertical relative humidity gradient. This research underscores the potential of deep learning in achieving accurate cloud parameterizations and emphasizes the effective role of symbolic regression in deriving interpretable, consistent equations for cloud cover. |
| abstractGer |
This thesis delves into the improvement of cloud parameterizations in climate models through machine learning trained on coarse-grained output from high-resolution simulations. Utilizing the ICOsahedral Non-hydrostatic (ICON) modeling framework, it specifically targets the enhancement of cloud cover parameterization within the ICON Earth System Model. Three types of neural networks (NNs) differing in vertical locality are developed to estimate cloud cover, with globally trained NNs even applicable to distinct regional simulations. Interpretability analysis exposes model-specific biases and local relationships with the thermodynamic environment. Despite achieving high predictive performance, NNs necessitate post-hoc interpretation tools. To tackle this issue, a combined hierarchical modeling framework incorporating symbolic regression, feature selection, and physical constraints is proposed. The resulting equations, characterized by simplicity and physical consistency, attain performance comparable to NNs while demonstrating superior transferability to other realistic datasets. Our best equation adeptly captures cloud cover distributions across various regimes, notably excelling in representing marine stratocumulus clouds by learning to utilize the vertical relative humidity gradient. This research underscores the potential of deep learning in achieving accurate cloud parameterizations and emphasizes the effective role of symbolic regression in deriving interpretable, consistent equations for cloud cover. |
| abstract_unstemmed |
This thesis delves into the improvement of cloud parameterizations in climate models through machine learning trained on coarse-grained output from high-resolution simulations. Utilizing the ICOsahedral Non-hydrostatic (ICON) modeling framework, it specifically targets the enhancement of cloud cover parameterization within the ICON Earth System Model. Three types of neural networks (NNs) differing in vertical locality are developed to estimate cloud cover, with globally trained NNs even applicable to distinct regional simulations. Interpretability analysis exposes model-specific biases and local relationships with the thermodynamic environment. Despite achieving high predictive performance, NNs necessitate post-hoc interpretation tools. To tackle this issue, a combined hierarchical modeling framework incorporating symbolic regression, feature selection, and physical constraints is proposed. The resulting equations, characterized by simplicity and physical consistency, attain performance comparable to NNs while demonstrating superior transferability to other realistic datasets. Our best equation adeptly captures cloud cover distributions across various regimes, notably excelling in representing marine stratocumulus clouds by learning to utilize the vertical relative humidity gradient. This research underscores the potential of deep learning in achieving accurate cloud parameterizations and emphasizes the effective role of symbolic regression in deriving interpretable, consistent equations for cloud cover. |
| collection_details |
GBV-ODiss GBV_ILN_20 ISIL_DE-84 SYSFLAG_1 GBV_KXP GBV_ILN_21 ISIL_DE-46 GBV_ILN_22 ISIL_DE-18 GBV_ILN_23 ISIL_DE-830 GBV_ILN_30 ISIL_DE-104 GBV_ILN_40 ISIL_DE-7 GBV_ILN_60 ISIL_DE-705 GBV_ILN_63 ISIL_DE-Wim2 GBV_ILN_70 ISIL_DE-89 GBV_ILN_105 ISIL_DE-841 GBV_ILN_110 ISIL_DE-Luen4 GBV_ILN_132 ISIL_DE-959 GBV_ILN_151 ISIL_DE-546 GBV_ILN_161 ISIL_DE-960 GBV_ILN_293 ISIL_DE-960-3 GBV_ILN_370 ISIL_DE-1373 GBV_ILN_2403 ISIL_DE-LFER |
| title_short |
Data-driven cloud cover parameterizations for the ICON earth system model using deep learning and symbolic regression |
| url |
https://doi.org/10.26092/elib/2821 https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397 https://d-nb.info/1339510391/34 https://media.suub.uni-bremen.de/handle/elib/7739 |
| ausleihindikator_str_mv |
20 21 22 23 30 40 60 63 70 105 110 132 151 161 293 370 2403:n |
| rolewithnormlink_str_mv |
@@655@@(DE-588)4113937-9 @@dgs@@(DE-588)1255681446 @@dgg@@(DE-588)2001386-3 @@uvp@@(DE-588)4008135-7 |
| remote_bool |
true |
| author2 |
Eyring, Veronika Gentine, Pierre Universität Bremen |
| author2Str |
Eyring, Veronika Gentine, Pierre Universität Bremen |
| ppnlink |
1883908809 |
| mediatype_str_mv |
c |
| isOA_txt |
true |
| hochschulschrift_bool |
true |
| author2_role |
dgs dgs |
| doi_str |
10.26092/elib/2821 |
| callnumber-a |
--%%-- |
| up_date |
2024-10-13T07:24:37.268Z |
| _version_ |
1812782673842667520 |
| fullrecord_marcxml |
<?xml version="1.0" encoding="UTF-8"?><collection xmlns="http://www.loc.gov/MARC21/slim"><record><leader>01000cam a2200265 4500</leader><controlfield tag="001">1883908477</controlfield><controlfield tag="003">DE-627</controlfield><controlfield tag="005">20240912032145.0</controlfield><controlfield tag="007">cr uuu---uuuuu</controlfield><controlfield tag="008">240320s2024 gw |||||om 00| ||eng c</controlfield><datafield tag="024" ind1="7" ind2=" "><subfield code="a">urn:nbn:de:gbv:46-elib77397</subfield><subfield code="2">urn</subfield></datafield><datafield tag="024" ind1="7" ind2=" "><subfield code="a">10.26092/elib/2821</subfield><subfield code="2">doi</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-627)1883908477</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-599)KXP1883908477</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(OCoLC)1427339102</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(OAEPFHB)elib/2821</subfield></datafield><datafield tag="040" ind1=" " ind2=" "><subfield code="a">DE-627</subfield><subfield code="b">ger</subfield><subfield code="c">DE-627</subfield><subfield code="e">rda</subfield></datafield><datafield tag="041" ind1=" " ind2=" "><subfield code="a">eng</subfield></datafield><datafield tag="044" ind1=" " ind2=" "><subfield code="c">XA-DE-HB</subfield></datafield><datafield tag="082" ind1="0" ind2="4"><subfield code="a">551.576</subfield><subfield code="q">DE-101</subfield></datafield><datafield tag="082" ind1="0" ind2="4"><subfield code="a">550</subfield><subfield code="q">DE-101</subfield></datafield><datafield tag="100" ind1="1" ind2=" "><subfield code="a">Grundner, Arthur</subfield><subfield code="e">verfasserin</subfield><subfield code="0">(orcid)0000-0002-3765-242X</subfield><subfield code="4">aut</subfield></datafield><datafield tag="245" ind1="1" ind2="0"><subfield code="a">Data-driven cloud cover parameterizations for the ICON earth system model using deep learning and symbolic regression</subfield><subfield code="c">Arthur Grundner</subfield></datafield><datafield tag="264" ind1=" " ind2="1"><subfield code="a">Bremen</subfield><subfield code="c">2024</subfield></datafield><datafield tag="300" ind1=" " ind2=" "><subfield code="a">1 Online-Ressource (xi, 140 Seiten)</subfield><subfield code="b">Illustrationen, Diagramme</subfield></datafield><datafield tag="336" ind1=" " ind2=" "><subfield code="a">Text</subfield><subfield code="b">txt</subfield><subfield code="2">rdacontent</subfield></datafield><datafield tag="337" ind1=" " ind2=" "><subfield code="a">Computermedien</subfield><subfield code="b">c</subfield><subfield code="2">rdamedia</subfield></datafield><datafield tag="338" ind1=" " ind2=" "><subfield code="a">Online-Ressource</subfield><subfield code="b">cr</subfield><subfield code="2">rdacarrier</subfield></datafield><datafield tag="502" ind1=" " ind2=" "><subfield code="b">Dissertation</subfield><subfield code="c">Universität Bremen</subfield><subfield code="d">2024</subfield></datafield><datafield tag="506" ind1="0" ind2=" "><subfield code="q">DE-46</subfield><subfield code="a">Open Access</subfield><subfield code="e">Controlled Vocabulary for Access Rights</subfield><subfield code="u">http://purl.org/coar/access_right/c_abf2</subfield></datafield><datafield tag="520" ind1=" " ind2=" "><subfield code="a">This thesis delves into the improvement of cloud parameterizations in climate models through machine learning trained on coarse-grained output from high-resolution simulations. Utilizing the ICOsahedral Non-hydrostatic (ICON) modeling framework, it specifically targets the enhancement of cloud cover parameterization within the ICON Earth System Model. Three types of neural networks (NNs) differing in vertical locality are developed to estimate cloud cover, with globally trained NNs even applicable to distinct regional simulations. Interpretability analysis exposes model-specific biases and local relationships with the thermodynamic environment. Despite achieving high predictive performance, NNs necessitate post-hoc interpretation tools. To tackle this issue, a combined hierarchical modeling framework incorporating symbolic regression, feature selection, and physical constraints is proposed. The resulting equations, characterized by simplicity and physical consistency, attain performance comparable to NNs while demonstrating superior transferability to other realistic datasets. Our best equation adeptly captures cloud cover distributions across various regimes, notably excelling in representing marine stratocumulus clouds by learning to utilize the vertical relative humidity gradient. This research underscores the potential of deep learning in achieving accurate cloud parameterizations and emphasizes the effective role of symbolic regression in deriving interpretable, consistent equations for cloud cover.</subfield></datafield><datafield tag="540" ind1=" " ind2=" "><subfield code="q">DE-46</subfield><subfield code="a">Namensnennung 4.0 International</subfield><subfield code="f">CC BY 4.0</subfield><subfield code="2">cc</subfield><subfield code="u">https://creativecommons.org/licenses/by/4.0/</subfield></datafield><datafield tag="583" ind1="1" ind2=" "><subfield code="a">Archivierung/Langzeitarchivierung gewährleistet</subfield><subfield code="f">PEHB</subfield><subfield code="x">XA-DE-HB</subfield><subfield code="2">pdager</subfield><subfield code="5">DE-46</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Cloud Cover</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Parameterization</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Deep Learning</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Neural Networks</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">ICON</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Pareto Frontier</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">PySR</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Sequential Feature Selection</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Equation Discovery</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Physical Constraints</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Machine Learning</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Symbolic Regression</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">SHAP</subfield></datafield><datafield tag="655" ind1=" " ind2="7"><subfield code="a">Hochschulschrift</subfield><subfield code="0">(DE-588)4113937-9</subfield><subfield code="0">(DE-627)105825778</subfield><subfield code="0">(DE-576)209480580</subfield><subfield code="2">gnd-content</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Eyring, Veronika</subfield><subfield code="e">akademischer betreuerin</subfield><subfield code="0">(DE-588)1255681446</subfield><subfield code="0">(DE-627)1799770168</subfield><subfield code="4">dgs</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Gentine, Pierre</subfield><subfield code="e">akademischer betreuerin</subfield><subfield code="4">dgs</subfield></datafield><datafield tag="710" ind1="2" ind2=" "><subfield code="a">Universität Bremen</subfield><subfield code="e">Grad-verleihende Institution</subfield><subfield code="0">(DE-588)2001386-3</subfield><subfield code="0">(DE-627)101380429</subfield><subfield code="0">(DE-576)191575038</subfield><subfield code="4">dgg</subfield></datafield><datafield tag="751" ind1=" " ind2=" "><subfield code="a">Bremen</subfield><subfield code="0">(DE-588)4008135-7</subfield><subfield code="0">(DE-627)106369636</subfield><subfield code="0">(DE-576)208874569</subfield><subfield code="4">uvp</subfield></datafield><datafield tag="776" ind1="0" ind2="8"><subfield code="i">Erscheint auch als</subfield><subfield code="n">Druck-Ausgabe</subfield><subfield code="a">Grundner, Arthur</subfield><subfield code="t">Data-driven cloud cover parameterizations for the ICON earth system model using deep learning and symbolic regression</subfield><subfield code="d">Bremen, 2023</subfield><subfield code="h">xi, 140 Seiten</subfield><subfield code="w">(DE-627)1883908809</subfield></datafield><datafield tag="856" ind1="4" ind2="0"><subfield code="u">https://doi.org/10.26092/elib/2821</subfield><subfield code="x">Resolving-System</subfield><subfield code="z">kostenfrei</subfield></datafield><datafield tag="856" ind1="4" ind2="0"><subfield code="u">https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397</subfield><subfield code="x">Resolving-System</subfield><subfield code="z">kostenfrei</subfield></datafield><datafield tag="856" ind1="4" ind2="0"><subfield code="u">https://d-nb.info/1339510391/34</subfield><subfield code="x">Langzeitarchivierung Nationalbibliothek</subfield><subfield code="z">kostenfrei</subfield></datafield><datafield tag="856" ind1="4" ind2="0"><subfield code="u">https://media.suub.uni-bremen.de/handle/elib/7739</subfield><subfield code="x">Verlag</subfield><subfield code="z">kostenfrei</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV-ODiss</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_20</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-84</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SYSFLAG_1</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_KXP</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_21</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-46</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_22</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-18</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_23</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-830</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_30</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-104</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_40</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-7</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_60</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-705</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_63</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-Wim2</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_70</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-89</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_105</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-841</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_110</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-Luen4</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_132</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-959</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_151</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-546</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_161</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-960</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_293</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-960-3</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_370</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-1373</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2403</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">ISIL_DE-LFER</subfield></datafield><datafield tag="935" ind1=" " ind2=" "><subfield code="i">DSpace</subfield></datafield><datafield tag="951" ind1=" " ind2=" "><subfield code="a">BO</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">20</subfield><subfield code="1">01</subfield><subfield code="x">0084</subfield><subfield code="b">4593082714</subfield><subfield code="y">x</subfield><subfield code="z">12-10-24</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">21</subfield><subfield code="1">01</subfield><subfield code="x">0046</subfield><subfield code="b">4502027553</subfield><subfield code="h">ebook_2024_dissbremen</subfield><subfield code="k">Kostenloser Zugriff</subfield><subfield code="y">zza</subfield><subfield code="z">20-03-24</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">22</subfield><subfield code="1">01</subfield><subfield code="x">0018</subfield><subfield code="b">4593184053</subfield><subfield code="h">SUBolrd</subfield><subfield code="y">xu</subfield><subfield code="z">12-10-24</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">23</subfield><subfield code="1">01</subfield><subfield code="x">0830</subfield><subfield code="b">459323218X</subfield><subfield code="h">olr-d</subfield><subfield code="y">x</subfield><subfield code="z">12-10-24</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">30</subfield><subfield code="1">01</subfield><subfield code="x">0104</subfield><subfield code="b">4593278708</subfield><subfield code="y">z</subfield><subfield code="z">12-10-24</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">40</subfield><subfield code="1">01</subfield><subfield code="x">0007</subfield><subfield code="b">4593314887</subfield><subfield code="y">xsn</subfield><subfield code="z">12-10-24</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">60</subfield><subfield code="1">01</subfield><subfield code="x">0705</subfield><subfield code="b">4593373239</subfield><subfield code="h">OLRD</subfield><subfield code="y">z</subfield><subfield code="z">12-10-24</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">63</subfield><subfield code="1">01</subfield><subfield code="x">3401</subfield><subfield code="b">4593426561</subfield><subfield code="h">ORD</subfield><subfield code="y">x</subfield><subfield code="z">12-10-24</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">70</subfield><subfield code="1">01</subfield><subfield code="x">0089</subfield><subfield code="b">4593481716</subfield><subfield code="y">z</subfield><subfield code="z">12-10-24</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">105</subfield><subfield code="1">01</subfield><subfield code="x">0841</subfield><subfield code="b">4593871697</subfield><subfield code="y">z</subfield><subfield code="z">12-10-24</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">110</subfield><subfield code="1">01</subfield><subfield code="x">3110</subfield><subfield code="b">4593582741</subfield><subfield code="y">x</subfield><subfield code="z">12-10-24</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">132</subfield><subfield code="1">01</subfield><subfield code="x">0959</subfield><subfield code="b">4593626374</subfield><subfield code="h">OLR-DISS</subfield><subfield code="y">x</subfield><subfield code="z">12-10-24</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">151</subfield><subfield code="1">01</subfield><subfield code="x">0546</subfield><subfield code="b">4593667224</subfield><subfield code="h">OLR-ODISS</subfield><subfield code="y">z</subfield><subfield code="z">12-10-24</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">161</subfield><subfield code="1">01</subfield><subfield code="x">0960</subfield><subfield code="b">4593691222</subfield><subfield code="h">ORD</subfield><subfield code="y">z</subfield><subfield code="z">12-10-24</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">293</subfield><subfield code="1">01</subfield><subfield code="x">3293</subfield><subfield code="b">4593820650</subfield><subfield code="h">ORD</subfield><subfield code="y">z</subfield><subfield code="z">12-10-24</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">370</subfield><subfield code="1">01</subfield><subfield code="x">4370</subfield><subfield code="b">4593854458</subfield><subfield code="y">x</subfield><subfield code="z">12-10-24</subfield></datafield><datafield tag="980" ind1=" " ind2=" "><subfield code="2">2403</subfield><subfield code="1">01</subfield><subfield code="x">DE-LFER</subfield><subfield code="b">4511167990</subfield><subfield code="c">00</subfield><subfield code="f">--%%--</subfield><subfield code="d">--%%--</subfield><subfield code="e">n</subfield><subfield code="j">--%%--</subfield><subfield code="y">l01</subfield><subfield code="z">12-04-24</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">20</subfield><subfield code="1">01</subfield><subfield code="x">0084</subfield><subfield code="r">https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">21</subfield><subfield code="1">01</subfield><subfield code="x">0046</subfield><subfield code="r">https://doi.org/10.26092/elib/2821</subfield><subfield code="z">LF</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">22</subfield><subfield code="1">01</subfield><subfield code="x">0018</subfield><subfield code="r">https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">23</subfield><subfield code="1">01</subfield><subfield code="x">0830</subfield><subfield code="r">https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">30</subfield><subfield code="1">01</subfield><subfield code="x">0104</subfield><subfield code="r">https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">40</subfield><subfield code="1">01</subfield><subfield code="x">0007</subfield><subfield code="r">https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">60</subfield><subfield code="1">01</subfield><subfield code="x">0705</subfield><subfield code="r">https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">63</subfield><subfield code="1">01</subfield><subfield code="x">3401</subfield><subfield code="y">E-Book</subfield><subfield code="r">https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397</subfield><subfield code="z">LF</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">70</subfield><subfield code="1">01</subfield><subfield code="x">0089</subfield><subfield code="r">https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">105</subfield><subfield code="1">01</subfield><subfield code="x">0841</subfield><subfield code="r">https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">110</subfield><subfield code="1">01</subfield><subfield code="x">3110</subfield><subfield code="r">https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">132</subfield><subfield code="1">01</subfield><subfield code="x">0959</subfield><subfield code="r">https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">151</subfield><subfield code="1">01</subfield><subfield code="x">0546</subfield><subfield code="y">Volltext</subfield><subfield code="r">https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">161</subfield><subfield code="1">01</subfield><subfield code="x">0960</subfield><subfield code="r">https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">293</subfield><subfield code="1">01</subfield><subfield code="x">3293</subfield><subfield code="r">https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">370</subfield><subfield code="1">01</subfield><subfield code="x">4370</subfield><subfield code="r">https://nbn-resolving.org/urn:nbn:de:gbv:46-elib77397</subfield></datafield><datafield tag="981" ind1=" " ind2=" "><subfield code="2">2403</subfield><subfield code="1">01</subfield><subfield code="x">DE-LFER</subfield><subfield code="r">https://doi.org/10.26092/elib/2821</subfield></datafield><datafield tag="982" ind1=" " ind2=" "><subfield code="2">21</subfield><subfield code="1">00</subfield><subfield code="x">DE-46</subfield><subfield code="8">00</subfield><subfield code="a">Universität Bremen</subfield></datafield><datafield tag="982" ind1=" " ind2=" "><subfield code="2">21</subfield><subfield code="1">00</subfield><subfield code="x">DE-46</subfield><subfield code="8">00</subfield><subfield code="a">Fachbereich 01: Physik/Elektrotechnik (FB 01)</subfield></datafield><datafield tag="983" ind1=" " ind2=" "><subfield code="2">60</subfield><subfield code="1">01</subfield><subfield code="x">0705</subfield><subfield code="8">10</subfield><subfield code="a">ho</subfield></datafield><datafield tag="985" ind1=" " ind2=" "><subfield code="2">20</subfield><subfield code="1">01</subfield><subfield code="x">0084</subfield><subfield code="a">OLRD</subfield></datafield><datafield tag="985" ind1=" " ind2=" "><subfield code="2">110</subfield><subfield code="1">01</subfield><subfield code="x">3110</subfield><subfield code="a">OLRD</subfield></datafield><datafield tag="985" ind1=" " ind2=" "><subfield code="2">370</subfield><subfield code="1">01</subfield><subfield code="x">4370</subfield><subfield code="a">OLRD</subfield></datafield><datafield tag="995" ind1=" " ind2=" "><subfield code="2">21</subfield><subfield code="1">01</subfield><subfield code="x">0046</subfield><subfield code="a">ebook_2024_dissbremen</subfield></datafield><datafield tag="995" ind1=" " ind2=" "><subfield code="2">22</subfield><subfield code="1">01</subfield><subfield code="x">0018</subfield><subfield code="a">SUBolrd</subfield></datafield><datafield tag="995" ind1=" " ind2=" "><subfield code="2">23</subfield><subfield code="1">01</subfield><subfield code="x">0830</subfield><subfield code="a">olr-d</subfield></datafield><datafield tag="995" ind1=" " ind2=" "><subfield code="2">60</subfield><subfield code="1">01</subfield><subfield code="x">0705</subfield><subfield code="a">OLRD</subfield></datafield><datafield tag="995" ind1=" " ind2=" "><subfield code="2">63</subfield><subfield code="1">01</subfield><subfield code="x">3401</subfield><subfield code="a">ORD</subfield></datafield><datafield tag="995" ind1=" " ind2=" "><subfield code="2">132</subfield><subfield code="1">01</subfield><subfield code="x">0959</subfield><subfield code="a">OLR-DISS</subfield></datafield><datafield tag="995" ind1=" " ind2=" "><subfield code="2">151</subfield><subfield code="1">01</subfield><subfield code="x">0546</subfield><subfield code="a">OLR-ODISS</subfield></datafield><datafield tag="995" ind1=" " ind2=" "><subfield code="2">161</subfield><subfield code="1">01</subfield><subfield code="x">0960</subfield><subfield code="a">ORD</subfield></datafield><datafield tag="995" ind1=" " ind2=" "><subfield code="2">293</subfield><subfield code="1">01</subfield><subfield code="x">3293</subfield><subfield code="a">ORD</subfield></datafield><datafield tag="998" ind1=" " ind2=" "><subfield code="2">23</subfield><subfield code="1">01</subfield><subfield code="x">0830</subfield><subfield code="0">2024-10-12 10:31:16</subfield></datafield></record></collection>
|
| score |
6.9003334 |