Unravelling the relationship between grey–green spatial structure and county-level carbon emission density across composite development typologies in the Yangtze River Delta

The transition towards urban carbon neutrality has intensified attention to the relationship between county-level carbon emission density (CED) and grey–green spatial structure across different development contexts. However, existing studies often examine socio-economic development, spatial growth, and landscape characteristics separately, limiting understanding of their joint relationships and nonlinear responses. This study examined 303 counties in the Yangtze River Delta in 2018 and 2023, measured CED within Urban Physical Areas, and constructed 18 composite typologies by integrating socio-economic development contexts with spatial growth morphologies. Six grey–green indicators were analysed using XGBoost–SHAP, contribution-pattern clustering, and nonlinear response analysis. CED did not follow a simple growth–shrinkage gradient but varied with combinations of development context and growth morphology. PLADJ ranked first in all 18 typologies in 2018 and remained the leading contributor in 14 typologies in 2023, while AREA_MN_Green ranked first in the remaining four. Hierarchical clustering identified three recurring contribution patterns. PLADJ showed a stable cross-group response threshold of approximately 97%, whereas AREA_MN_Green and LSI_Grey showed more group-specific responses. This study provides an interpretable framework linking typology-specific contribution patterns and nonlinear responses to differentiated low-carbon spatial planning.