Dice Question Streamline Icon: https://streamlinehq.com

Including external covariates in CTRW and BSAR models

Extend the Coupled Continuous Time Random Walk (CTRW) model and the Bayesian Spectral Analysis Regression (BSAR) model with an isotonic Gaussian Process prior to include external covariates such as CO2 emissions, aerosol concentrations, and urbanization indices in order to attribute observed annual maximum temperature dynamics to anthropogenic and natural drivers.

Information Square Streamline Icon: https://streamlinehq.com

Background

The paper compares CTRW and BSAR models using time as the primary predictor of annual maximum temperature trends. For attribution and improved explanatory power, the authors highlight the importance of incorporating external covariates reflecting anthropogenic and natural influences.

They explicitly mark this direction as an open avenue, proposing the inclusion of covariates such as CO2 emissions, aerosols, and urbanization indices.

References

Several avenues remain open for further exploration. Second, extending the CTRW and BSAR models to include external covariates (e.g., CO$_2$ emissions, aerosol concentrations, urbanization indices) can help attribute observed temperature dynamics to anthropogenic and natural factors.

Bayesian Modeling of Long-Term Dynamics in Indian Temperature Extremes (2507.01540 - Chakraborty, 2 Jul 2025) in Section 5 (Conclusion)