Benchmarking of Data-Driven Causality Discovery Approaches in the Interactions of Arctic Sea Ice and Atmosphere
Copyright © 2021 Huang, Kleindessner, Munishkin, Varshney, Guo and Wang..
The Arctic sea ice has retreated rapidly in the past few decades, which is believed to be driven by various dynamic and thermodynamic processes in the atmosphere. The newly open water resulted from sea ice decline in turn exerts large influence on the atmosphere. Therefore, this study aims to investigate the causality between multiple atmospheric processes and sea ice variations using three distinct data-driven causality approaches that have been proposed recently: Temporal Causality Discovery Framework Non-combinatorial Optimization via Trace Exponential and Augmented lagrangian for Structure learning (NOTEARS) and Directed Acyclic Graph-Graph Neural Networks (DAG-GNN). We apply these three algorithms to 39 years of historical time-series data sets, which include 11 atmospheric variables from ERA-5 reanalysis product and passive microwave satellite retrieved sea ice extent. By comparing the causality graph results of these approaches with what we summarized from the literature, it shows that the static graphs produced by NOTEARS and DAG-GNN are relatively reasonable. The results from NOTEARS indicate that relative humidity and precipitation dominate sea ice changes among all variables, while the results from DAG-GNN suggest that the horizontal and meridional wind are more important for driving sea ice variations. However, both approaches produce some unrealistic cause-effect relationships. Additionally, these three methods cannot well detect the delayed impact of one variable on another in the Arctic. It also turns out that the results are rather sensitive to the choice of hyperparameters of the three methods. As a pioneer study, this work paves the way to disentangle the complex causal relationships in the Earth system, by taking the advantage of cutting-edge Artificial Intelligence technologies.
Medienart: |
E-Artikel |
---|
Erscheinungsjahr: |
2021 |
---|---|
Erschienen: |
2021 |
Enthalten in: |
Zur Gesamtaufnahme - volume:4 |
---|---|
Enthalten in: |
Frontiers in big data - 4(2021) vom: 01., Seite 642182 |
Sprache: |
Englisch |
---|
Beteiligte Personen: |
Huang, Yiyi [VerfasserIn] |
---|
Links: |
---|
Anmerkungen: |
Date Revised 11.09.2021 published: Electronic-eCollection Citation Status PubMed-not-MEDLINE |
---|
doi: |
10.3389/fdata.2021.642182 |
---|
funding: |
|
---|---|
Förderinstitution / Projekttitel: |
|
PPN (Katalog-ID): |
NLM33047295X |
---|
LEADER | 01000naa a22002652 4500 | ||
---|---|---|---|
001 | NLM33047295X | ||
003 | DE-627 | ||
005 | 20231225211346.0 | ||
007 | cr uuu---uuuuu | ||
008 | 231225s2021 xx |||||o 00| ||eng c | ||
024 | 7 | |a 10.3389/fdata.2021.642182 |2 doi | |
028 | 5 | 2 | |a pubmed24n1101.xml |
035 | |a (DE-627)NLM33047295X | ||
035 | |a (NLM)34505056 | ||
040 | |a DE-627 |b ger |c DE-627 |e rakwb | ||
041 | |a eng | ||
100 | 1 | |a Huang, Yiyi |e verfasserin |4 aut | |
245 | 1 | 0 | |a Benchmarking of Data-Driven Causality Discovery Approaches in the Interactions of Arctic Sea Ice and Atmosphere |
264 | 1 | |c 2021 | |
336 | |a Text |b txt |2 rdacontent | ||
337 | |a ƒaComputermedien |b c |2 rdamedia | ||
338 | |a ƒa Online-Ressource |b cr |2 rdacarrier | ||
500 | |a Date Revised 11.09.2021 | ||
500 | |a published: Electronic-eCollection | ||
500 | |a Citation Status PubMed-not-MEDLINE | ||
520 | |a Copyright © 2021 Huang, Kleindessner, Munishkin, Varshney, Guo and Wang. | ||
520 | |a The Arctic sea ice has retreated rapidly in the past few decades, which is believed to be driven by various dynamic and thermodynamic processes in the atmosphere. The newly open water resulted from sea ice decline in turn exerts large influence on the atmosphere. Therefore, this study aims to investigate the causality between multiple atmospheric processes and sea ice variations using three distinct data-driven causality approaches that have been proposed recently: Temporal Causality Discovery Framework Non-combinatorial Optimization via Trace Exponential and Augmented lagrangian for Structure learning (NOTEARS) and Directed Acyclic Graph-Graph Neural Networks (DAG-GNN). We apply these three algorithms to 39 years of historical time-series data sets, which include 11 atmospheric variables from ERA-5 reanalysis product and passive microwave satellite retrieved sea ice extent. By comparing the causality graph results of these approaches with what we summarized from the literature, it shows that the static graphs produced by NOTEARS and DAG-GNN are relatively reasonable. The results from NOTEARS indicate that relative humidity and precipitation dominate sea ice changes among all variables, while the results from DAG-GNN suggest that the horizontal and meridional wind are more important for driving sea ice variations. However, both approaches produce some unrealistic cause-effect relationships. Additionally, these three methods cannot well detect the delayed impact of one variable on another in the Arctic. It also turns out that the results are rather sensitive to the choice of hyperparameters of the three methods. As a pioneer study, this work paves the way to disentangle the complex causal relationships in the Earth system, by taking the advantage of cutting-edge Artificial Intelligence technologies | ||
650 | 4 | |a Journal Article | |
650 | 4 | |a arctic sea ice | |
650 | 4 | |a atmosphere-sea ice interactions | |
650 | 4 | |a causality discovery | |
650 | 4 | |a directed acyclic graph-graph neural networks | |
650 | 4 | |a non-combinatorial optimization via trace exponential and augmented lagrangian for structure learning | |
650 | 4 | |a temporal causality discovery framework | |
650 | 4 | |a time series | |
700 | 1 | |a Kleindessner, Matthäus |e verfasserin |4 aut | |
700 | 1 | |a Munishkin, Alexey |e verfasserin |4 aut | |
700 | 1 | |a Varshney, Debvrat |e verfasserin |4 aut | |
700 | 1 | |a Guo, Pei |e verfasserin |4 aut | |
700 | 1 | |a Wang, Jianwu |e verfasserin |4 aut | |
773 | 0 | 8 | |i Enthalten in |t Frontiers in big data |d 2018 |g 4(2021) vom: 01., Seite 642182 |w (DE-627)NLM315255927 |x 2624-909X |7 nnns |
773 | 1 | 8 | |g volume:4 |g year:2021 |g day:01 |g pages:642182 |
856 | 4 | 0 | |u http://dx.doi.org/10.3389/fdata.2021.642182 |3 Volltext |
912 | |a GBV_USEFLAG_A | ||
912 | |a GBV_NLM | ||
951 | |a AR | ||
952 | |d 4 |j 2021 |b 01 |h 642182 |