City-level carbon emission data in Southeast Asia from 2000 to 2020

This study presents a city-level carbon emissions dataset for SEA from 2000 to 2020, constructed within a top‑down allocation framework that anchors city estimates to EDGAR provincial totals. By integrating multiple proxies: impervious surfaces, nighttime lights, and urban–rural residential distributions, and evaluating Random Forest, Gradient Boosting, and Neural Network algorithms within an adaptive dual‑modeling strategy, the approach delivers a consistent, two‑decade panel for 4,413 city‑level units. This work was supported by the Lee Kuan Yew School of Public Policy, National University of Singapore.