The abstract is read at the publisher; the summary is JClub's.
Pixel-scale SSA-ConvLSTM best predicts carbon emissions from energy consumption, while Ister-Transformer is optimal at county and city scales in the Pearl River Delta from 2000–2022.
- Why it matters: Regional low-carbon planning needs more than regression or trend extrapolation; it requires multi-scale causal mechanisms and tested multi-scenario pathways for carbon emissions driven by energy use.
- What they did: The study integrates multi-source socioeconomic, land-use, and climate data for the Pearl River Delta urban agglomeration (2000–2022), using SHAP for key-variable identification, causal graphs for net effects and transmission pathways, and sub-scale optimal models for scenario prediction.
- The result: Causal inference links pixel-scale PGDP to lower emissions, with county-scale GDP and JCQ acting mainly via population and city-scale population showing the strongest direct effect; stratified policy scenarios are simulated by municipal, county, and pixel scales.
The findingWhy it mattersWhat they didThe result
AI summary of the abstract · may contain errors · not medical advice · read the paper.