Abstract
The contradiction between rapid urbanization and ecological protection leads to the aggravation of carbon sink function differentiation in coastal areas. In this study, a framework of "historical assessment-future prediction-mechanism analysis" was constructed, and InVEST, PLUS, SEM, and OPGD models were integrated to study the dynamics of carbon storage and its driving mechanisms in three rapidly urbanizing coastal areas (Beidaihe, Caofeidian, and Huanghua) in the Bohai Rim region of China. The main results showed that: (1) carbon storage showed the characteristics of periodic change—the expansion of low-carbon land led to the decline of carbon storage (2000-2010), and then Caofeidian achieved $1.50 \times 10^6$ tons of carbon storage restoration through ecological restoration (2010-2020). (2) Carbon stocks are expected to increase in all four scenarios in 2030, with the largest increase in the Cultivated Land Protection (CP) scenario, with a peak of $3.00 \times 10^7$ tons. (3) The analysis of the driving mechanism showed that population density was the main driving factor ($q = 0.2126–0.2129$), and there was a significant nonlinear interaction with socio-economic factors, which jointly shaped the spatial pattern. This analysis reveals the spatiotemporal heterogeneity and core driving factors of carbon storage changes in coastal areas, emphasizes the key role of human activities under urbanization and ecological conflicts, and provides a scientific basis for the implementation of differentiated low-carbon spatial planning in rapidly developing coastal areas.
Introduction
Coastal zones are ecological "double agents": they are incredibly efficient carbon sinks but also the primary frontiers for rapid urbanization. In the Bohai Rim region of China, the tension between economic expansion and ecological preservation has led to significant fluctuations in carbon storage. Existing studies often struggle to bridge the gap between mapping land-use change and identifying complex causal drivers. This study addresses this gap by investigating three distinct coastal archetypes: Beidaihe (tourism), Caofeidian (heavy industry/port), and Huanghua (saline-alkali agriculture).
Study Area and Methodology
The research focuses on three key districts in Hebei Province, China, situated along the Bohai Sea. These areas represent diverse development paths and ecological challenges.
To analyze these regions, the authors constructed a "historical assessment-future prediction-mechanism analysis" framework using four integrated models:
- InVEST (Integrated Valuation of Ecosystem Services and Tradeoffs): Quantified carbon storage across four pools (aboveground, belowground, soil, and dead organic matter).
- PLUS (Patch-generating Land Use Simulation): Simulated land-use changes for 2030 across four scenarios: Natural Development (ND), Cultivated Land Protection (CP), Economic Development (ED), and Ecological Protection (EP).
- OPGD (Optimal Parameter Geodetector): Identified which independent variables (e.g., GDP, slope) best explain the spatial heterogeneity of carbon stocks.
- SEM (Structural Equation Modeling): Identified direct and indirect causal paths between driving factors, moving beyond simple correlation.
Key Insights and Results
Historical Trends and Periodic Change
Carbon storage in the region exhibited "periodic change" rather than a linear decline. Between 2000 and 2010, urban sprawl caused a sharp drop in carbon sinks. However, from 2010 to 2020, Caofeidian successfully restored $1.50 \times 10^6$ tons of carbon storage through coastal wetland restoration, demonstrating that targeted human intervention can effectively offset urbanization losses. Overall, Beidaihe and Huanghua saw net carbon reductions of $2.35 \times 10^5$ and $6.37 \times 10^5$ tons respectively from 2000–2020, while Caofeidian saw a net increase of $1.21 \times 10^6$ tons.
2030 Projections and the CP Scenario
Carbon stocks are expected to rise across all simulated scenarios by 2030. Notably, the Cultivated Land Protection (CP) scenario—rather than the Ecological Protection (EP) scenario—yielded the highest total carbon stock increase, peaking at approximately 30 million tons ($3.00 \times 10^7$). This highlights the significant carbon-sequestration potential of maintaining stable, high-quality agricultural soils in coastal regions.
Driving Mechanisms
Population density was identified as the dominant driver ($q \approx 0.21$). However, it does not act in isolation; it exhibits a significant nonlinear interaction with GDP and transport factors. The impact of a growing population on carbon loss is amplified when combined with high economic density. SEM analysis further confirmed that human activities have a direct negative impact on carbon stocks, often mediated by climate factors such as temperature and precipitation.
Discussion and Limitations
While the integrated model provides a robust framework for coastal planning, several limitations remain. The InVEST model utilizes static carbon densities, which may overlook the dynamic nature of vegetation succession and tidal inundation in "blue carbon" ecosystems. Additionally, the SEM model suggested that the 13 selected drivers do not capture the full complexity of the ecosystem, indicating a need for more granular data.
Future research should incorporate Monte Carlo simulations to improve causal credibility and utilize customized land-use maps to better capture small-scale variations in coastal ecosystems. This study provides a vital scientific roadmap for achieving China's "dual carbon" goals—peaking emissions by 2030 and achieving neutrality by 2060—within rapidly developing coastal zones.
Summary updated