Fine-scale urban-land configurations indirectly drive the strong spatial variability in pCO2 and CO2 fluxes in mountainous urban streams.

Urban streams are increasingly recognized as significant sources of atmospheric CO2. However, the spatial variability of CO2 fluxes in polluted mountainous urban streams, and how watershed-scale urban functional land use configurations indirectly regulate this variability, remain poorly constrained. This study conducted high-resolution spatial and seasonal monitoring of CO2 partial pressure (pCO2) and fluxes (fCO2) in two heavily polluted urban streams in Chongqing, China. We integrated a novel Environmental Pressure Index (EPI) based on fine-scale urban-land configurations to explore how urban environmental pressures drive the variability in stream CO2 emissions. The results showed that both streams were hotspots of CO2 emissions, with the overall mean of fCO2 (400 ± 387 mmol m-2 d-1) fivefold greater than the global mean for fluvial systems. Notably, fCO2 exhibited extreme spatial heterogeneity (3 to 1824 mmol m-2 d-1), varying by up to ∼400-fold over short stream reaches. This pronounced spatial variability was strongly linked to the fine-scale configurations of urban functional land-use, with fCO2 in high-EPI segments being up to 21.5 times greater than in low-EPI reaches. We identified two regulatory modes of EPI on fCO2: localized dominance due to high-intensity point sources (explaining 71% of variance in one stream) and longitudinal accumulation of upstream pressures (explaining over 90% in the other). We highlight EPI as effective predictors of the spatial variability in CO2 emissions from mountainous urban streams. Structural equation modeling further indicated that the urban land-use based EPI regulated fCO2 indirectly by governing in-stream water quality, with nitrogen-related variables emerging as dominant proximate catalyst for enhanced fCO2. Temperature and hydrology modulated the seasonal variation in CO2 fluxes, and the influence pathways of EPI on spatial heterogeneity varied seasonally. This study demonstrates that fine-scale urban land-use configurations and cumulative effects are fundamental, albeit indirect, controllers shaping the spatiotemporal variability of CO2 fluxes in heavily polluted streams. Incorporating such fine-scale anthropogenic landscape factors, as exemplified by the EPI framework, may be critical for predicting CO2 emission hotspot and accurately assessing carbon emission inventory from these urbanized waters.