Benchmarking of Data-Driven Causality Discovery Approaches in the Interactions of Arctic Sea Ice and Atmosphere

Abstract

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.

IceTFT v 1.0.0: Interpretable Long-Term Prediction of Arctic Sea Ice Extent with Deep Learning

“…Numerous studies have analyzed the causal relationship between sea ice and physical variables, due to the fact that fluctuations in sea ice can be generated by different dynamical and thermodynamical processes and other factors. Huang et al (2021) summarized recent studies and well-known atmospheric processes which connected with sea ice, and presented the causality graph as Fig. 3.…”

“…From the study of Huang et al (2021), the arrows b and c indicate that the increase of cloudiness and water vapor in the Arctic basin is due to local evaporation or enhanced water vapor transport, resulting in an increase in downward longwave radiation flux (DLWRF) (Luo et al, 2017). And DLWRF dominates surface warming and enhances sea ice melting in winter and spring (Kapsch et al, 2016(Kapsch et al, , 2013.…”

“…Higher precipitation and snowfall may lead to a thicker snowpack, which has implications for sea ice changing (Bintanja and Selten, 2014). Some researchers have studied (Huang et al, 2021) between important atmospheric variables and sea ice over the Arctic. Note that the processes a-d are well-known atmospheric processes, which may be outlined in multiple textbooks.…”

Causal discovery of drivers of surface ozone variability in Antarctica using a deep learning algorithm

“…We use one such algorithm called PCMCI, which extends the PC algorithm (named aer Peter Spirtes and Clark Glymour) to determine the true causal relationships and is the most widely used algorithm for causal discovery in climate sciences. 12,[32][33][34][35][36][37] The PCMCI algorithm consists of two steps:…”

Benchmarking Probabilistic Machine Learning Models for Arctic Sea Ice Forecasting

“…These records include monthly mean values of sea-ice extent (SIE) from Nimbus-7 SSMR and DMSP SSM/I-SSMIS passive microwave data version 2 and 9 meteorological data variables obtained from ERA-5 global reanalysis product. 3 The choice and details of these variables is presented in our previous causal discovery study conducted on Arctic sea-ice [6]. To conduct our experiments, we first combined all the raw variable datasets to have single temporal and spatial resolution, i.e.…”