Applying machine learning for drought prediction in a perfect model framework using data from a large ensemble of climate simulations
There is a strong scientific and social interest in understanding the factors leading to extreme events in order to improve the management of risks associated with hazards like droughts. In this study, artificial neural networks are applied to predict the occurrence of a drought in two contrasting E...
Published in: | Natural Hazards and Earth System Sciences |
---|---|
Main Authors: | , |
Format: | Article in Journal/Newspaper |
Language: | English |
Published: |
Copernicus Publications
2021
|
Subjects: | |
Online Access: | https://doi.org/10.5194/nhess-21-3679-2021 https://doaj.org/article/08f07a6ac58d472cbee517d11574ce75 |
id |
ftdoajarticles:oai:doaj.org/article:08f07a6ac58d472cbee517d11574ce75 |
---|---|
record_format |
openpolar |
spelling |
ftdoajarticles:oai:doaj.org/article:08f07a6ac58d472cbee517d11574ce75 2023-05-15T17:34:23+02:00 Applying machine learning for drought prediction in a perfect model framework using data from a large ensemble of climate simulations E. Felsche R. Ludwig 2021-12-01T00:00:00Z https://doi.org/10.5194/nhess-21-3679-2021 https://doaj.org/article/08f07a6ac58d472cbee517d11574ce75 EN eng Copernicus Publications https://nhess.copernicus.org/articles/21/3679/2021/nhess-21-3679-2021.pdf https://doaj.org/toc/1561-8633 https://doaj.org/toc/1684-9981 doi:10.5194/nhess-21-3679-2021 1561-8633 1684-9981 https://doaj.org/article/08f07a6ac58d472cbee517d11574ce75 Natural Hazards and Earth System Sciences, Vol 21, Pp 3679-3691 (2021) Environmental technology. Sanitary engineering TD1-1066 Geography. Anthropology. Recreation G Environmental sciences GE1-350 Geology QE1-996.5 article 2021 ftdoajarticles https://doi.org/10.5194/nhess-21-3679-2021 2022-12-31T13:02:24Z There is a strong scientific and social interest in understanding the factors leading to extreme events in order to improve the management of risks associated with hazards like droughts. In this study, artificial neural networks are applied to predict the occurrence of a drought in two contrasting European domains, Munich and Lisbon, with a lead time of 1 month. The approach takes into account a list of 28 atmospheric and soil variables as input parameters from a single-model initial-condition large ensemble (CRCM5-LE). The data were produced in the context of the ClimEx project by Ouranos, with the Canadian Regional Climate Model (CRCM5) driven by 50 members of the Canadian Earth System Model (CanESM2). Drought occurrence is defined using the standardized precipitation index. The best-performing machine learning algorithms manage to obtain a correct classification of drought or no drought for a lead time of 1 month for around 55 %–57 % of the events of each class for both domains. Explainable AI methods like SHapley Additive exPlanations (SHAP) are applied to understand the trained algorithms better. Variables like the North Atlantic Oscillation index and air pressure 1 month before the event prove essential for the prediction. The study shows that seasonality strongly influences the performance of drought prediction, especially for the Lisbon domain. Article in Journal/Newspaper North Atlantic North Atlantic oscillation Directory of Open Access Journals: DOAJ Articles Natural Hazards and Earth System Sciences 21 12 3679 3691 |
institution |
Open Polar |
collection |
Directory of Open Access Journals: DOAJ Articles |
op_collection_id |
ftdoajarticles |
language |
English |
topic |
Environmental technology. Sanitary engineering TD1-1066 Geography. Anthropology. Recreation G Environmental sciences GE1-350 Geology QE1-996.5 |
spellingShingle |
Environmental technology. Sanitary engineering TD1-1066 Geography. Anthropology. Recreation G Environmental sciences GE1-350 Geology QE1-996.5 E. Felsche R. Ludwig Applying machine learning for drought prediction in a perfect model framework using data from a large ensemble of climate simulations |
topic_facet |
Environmental technology. Sanitary engineering TD1-1066 Geography. Anthropology. Recreation G Environmental sciences GE1-350 Geology QE1-996.5 |
description |
There is a strong scientific and social interest in understanding the factors leading to extreme events in order to improve the management of risks associated with hazards like droughts. In this study, artificial neural networks are applied to predict the occurrence of a drought in two contrasting European domains, Munich and Lisbon, with a lead time of 1 month. The approach takes into account a list of 28 atmospheric and soil variables as input parameters from a single-model initial-condition large ensemble (CRCM5-LE). The data were produced in the context of the ClimEx project by Ouranos, with the Canadian Regional Climate Model (CRCM5) driven by 50 members of the Canadian Earth System Model (CanESM2). Drought occurrence is defined using the standardized precipitation index. The best-performing machine learning algorithms manage to obtain a correct classification of drought or no drought for a lead time of 1 month for around 55 %–57 % of the events of each class for both domains. Explainable AI methods like SHapley Additive exPlanations (SHAP) are applied to understand the trained algorithms better. Variables like the North Atlantic Oscillation index and air pressure 1 month before the event prove essential for the prediction. The study shows that seasonality strongly influences the performance of drought prediction, especially for the Lisbon domain. |
format |
Article in Journal/Newspaper |
author |
E. Felsche R. Ludwig |
author_facet |
E. Felsche R. Ludwig |
author_sort |
E. Felsche |
title |
Applying machine learning for drought prediction in a perfect model framework using data from a large ensemble of climate simulations |
title_short |
Applying machine learning for drought prediction in a perfect model framework using data from a large ensemble of climate simulations |
title_full |
Applying machine learning for drought prediction in a perfect model framework using data from a large ensemble of climate simulations |
title_fullStr |
Applying machine learning for drought prediction in a perfect model framework using data from a large ensemble of climate simulations |
title_full_unstemmed |
Applying machine learning for drought prediction in a perfect model framework using data from a large ensemble of climate simulations |
title_sort |
applying machine learning for drought prediction in a perfect model framework using data from a large ensemble of climate simulations |
publisher |
Copernicus Publications |
publishDate |
2021 |
url |
https://doi.org/10.5194/nhess-21-3679-2021 https://doaj.org/article/08f07a6ac58d472cbee517d11574ce75 |
genre |
North Atlantic North Atlantic oscillation |
genre_facet |
North Atlantic North Atlantic oscillation |
op_source |
Natural Hazards and Earth System Sciences, Vol 21, Pp 3679-3691 (2021) |
op_relation |
https://nhess.copernicus.org/articles/21/3679/2021/nhess-21-3679-2021.pdf https://doaj.org/toc/1561-8633 https://doaj.org/toc/1684-9981 doi:10.5194/nhess-21-3679-2021 1561-8633 1684-9981 https://doaj.org/article/08f07a6ac58d472cbee517d11574ce75 |
op_doi |
https://doi.org/10.5194/nhess-21-3679-2021 |
container_title |
Natural Hazards and Earth System Sciences |
container_volume |
21 |
container_issue |
12 |
container_start_page |
3679 |
op_container_end_page |
3691 |
_version_ |
1766133206839459840 |