Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Time-Varying Causal Treatment for Quantifying the Causal Effect of Short-Term Variations on Arctic Sea Ice Dynamics

Akila Sampath, Vandana Janeja, Jianwu Wang

TL;DR

The paper addresses the challenge of estimating causal effects in complex Arctic ocean–ice systems, where short-term SSH variations influence sea ice thickness through physical processes and unobserved confounding. It introduces KGCM-VAE, a knowledge-guided variational autoencoder with a causal-adjacency decoder and MMD-based latent deconfounding to learn balanced, physically-consistent representations for time-varying treatments. The approach leverages hydrostatic and geostrophic physics to guide treatment generation and enforces a learned causal structure, yielding superior PEHE performance over state-of-the-art baselines and demonstrating robustness across treatment lags. The work advances data-driven causal understanding in polar climate dynamics, enabling more reliable counterfactual analysis and informing climate model refinement and prediction under real-world observational constraints.

Abstract

Quantifying the causal relationship between ice melt and freshwater distribution is critical, as these complex interactions manifest as regional fluctuations in sea surface height (SSH). Leveraging SSH as a proxy for sea ice dynamics enables improved understanding of the feedback mechanisms driving polar climate change and global sea-level rise. However, conventional deep learning models often struggle with reliable treatment effect estimation in spatiotemporal settings due to unobserved confounders and the absence of physical constraints. To address these challenges, we propose the Knowledge-Guided Causal Model Variational Autoencoder (KGCM-VAE) to quantify causal mechanisms between sea ice thickness and SSH. The proposed framework integrates a velocity modulation scheme in which smoothed velocity signals are dynamically amplified via a sigmoid function governed by SSH transitions to generate physically grounded causal treatments. In addition, the model incorporates Maximum Mean Discrepancy (MMD) to balance treated and control covariate distributions in the latent space, along with a causal adjacency-constrained decoder to ensure alignment with established physical structures. Experimental results on both synthetic and real-world Arctic datasets demonstrate that KGCM-VAE achieves superior PEHE compared to state-of-the-art benchmarks. Ablation studies further confirm the effectiveness of the approach, showing that the joint application of MMD and causal adjacency constraints yields a 1.88\% reduction in estimation error.

Time-Varying Causal Treatment for Quantifying the Causal Effect of Short-Term Variations on Arctic Sea Ice Dynamics

TL;DR

The paper addresses the challenge of estimating causal effects in complex Arctic ocean–ice systems, where short-term SSH variations influence sea ice thickness through physical processes and unobserved confounding. It introduces KGCM-VAE, a knowledge-guided variational autoencoder with a causal-adjacency decoder and MMD-based latent deconfounding to learn balanced, physically-consistent representations for time-varying treatments. The approach leverages hydrostatic and geostrophic physics to guide treatment generation and enforces a learned causal structure, yielding superior PEHE performance over state-of-the-art baselines and demonstrating robustness across treatment lags. The work advances data-driven causal understanding in polar climate dynamics, enabling more reliable counterfactual analysis and informing climate model refinement and prediction under real-world observational constraints.

Abstract

Quantifying the causal relationship between ice melt and freshwater distribution is critical, as these complex interactions manifest as regional fluctuations in sea surface height (SSH). Leveraging SSH as a proxy for sea ice dynamics enables improved understanding of the feedback mechanisms driving polar climate change and global sea-level rise. However, conventional deep learning models often struggle with reliable treatment effect estimation in spatiotemporal settings due to unobserved confounders and the absence of physical constraints. To address these challenges, we propose the Knowledge-Guided Causal Model Variational Autoencoder (KGCM-VAE) to quantify causal mechanisms between sea ice thickness and SSH. The proposed framework integrates a velocity modulation scheme in which smoothed velocity signals are dynamically amplified via a sigmoid function governed by SSH transitions to generate physically grounded causal treatments. In addition, the model incorporates Maximum Mean Discrepancy (MMD) to balance treated and control covariate distributions in the latent space, along with a causal adjacency-constrained decoder to ensure alignment with established physical structures. Experimental results on both synthetic and real-world Arctic datasets demonstrate that KGCM-VAE achieves superior PEHE compared to state-of-the-art benchmarks. Ablation studies further confirm the effectiveness of the approach, showing that the joint application of MMD and causal adjacency constraints yields a 1.88\% reduction in estimation error.
Paper Structure (26 sections, 17 equations, 2 figures, 3 tables)

This paper contains 26 sections, 17 equations, 2 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Data and Problem Formulation
  4. Data
  5. Problem Formulation
  6. Methodology
  7. Knowledge Guidance
  8. Knowledge-Guided treatment generation
  9. Synthetic data generation
  10. Model Architecture
  11. Knowledge-Guided Causal Modeling (KGCM-VAE) Framework:
  12. Encoder and Latent Representation:
  13. Causal Decoder (CM):
  14. Causal Connection Matrix for Temporal Dependency:
  15. Temporal Discretization and Delayed Treatment Effects:
  16. ...and 11 more sections

Figures (2)

  • Figure 1: The Architecture of the Knowledge-Guided Causal Model Variational Autoencoder (KGCM-VAE) integrates a balanced latent space $\mathbf{z}$ learned by the encoder with a knowledge-guided causal model in the decoder for robust counterfactual prediction and Individual Treatment Effect estimation.
  • Figure 2: Real-world spatial analyses of factual and counterfactual outcomes. The panels display the temporal patterns of perturbed Ocean Velocity (Top left) and the corresponding counterfactual prediction of Sea Ice Thickness (Bottom left), alongside perturbed Sea Surface Height (Top right) and the corresponding counterfactual prediction of Sea Ice Thickness (Bottom right).