Generating Fine-Grained Causality in Climate Time Series Data for Forecasting and Anomaly Detection

TL;DR

This paper introduces a novel deep generative model, TacSas, that uncovers fine-grained causal relations in climate time series data, improving forecasting and anomaly detection by integrating Bayesian Networks with neural causality models.

Contribution

It proposes a new conceptual causal model TBN Granger Causality and an end-to-end deep generative framework TacSas for causal discovery, forecasting, and anomaly detection in complex climate data.

Findings

TacSas outperforms baseline models on Lorenz-96 causality benchmark.

TacSas improves climate forecasting accuracy on ERA5 dataset.

TacSas effectively detects anomalies in NOAA extreme weather data.

Abstract

Understanding the causal interaction of time series variables can contribute to time series data analysis for many real-world applications, such as climate forecasting and extreme weather alerts. However, causal relationships are difficult to be fully observed in real-world complex settings, such as spatial-temporal data from deployed sensor networks. Therefore, to capture fine-grained causal relations among spatial-temporal variables for further a more accurate and reliable time series analysis, we first design a conceptual fine-grained causal model named TBN Granger Causality, which adds time-respecting Bayesian Networks to the previous time-lagged Neural Granger Causality to offset the instantaneous effects. Second, we propose an end-to-end deep generative model called TacSas, which discovers TBN Granger Causality in a generative manner to help forecast time series data and detect possible anomalies during the forecast. For evaluations, besides the causality discovery benchmark Lorenz-96, we also test TacSas on climate benchmark ERA5 for climate forecasting and the extreme weather benchmark of NOAA for extreme weather alerts.