Cohesive soil in the reservoir area is vulnerable to natural disasters because of its poor erosion resistance and low strength. Therefore, it needs to be reinforced. Microbially induced calcium carbonate precipitation (MICP) is a sustaibable soil reinforcement technique with low energy consumption and no pollution. Different combinations of Bacillus subtilis bacterial solution (BS) concentrations and cementing solution (CS) concentrations were set to perform MICP solidification treatment. The characterization of cohesive soil before MICP was carried out by means of Scanning Electron Microscopy (SEM), Fourier-Transform Infrared Spectroscopy (FTIR), and Laser Particle Size Analyzer (LPSA). The results showed that the unreinforced soil showed an amorphous state with low strength and the particle size distribution was dominated by powder particles. However, with the addition of BS concentrations and CS concentrations, SEM results showed that spherical and rhombohedral minerals filled the pores of the cohesive soil, which increased the content of precipitations and enhanced the cementitious characteristics. When the concentrations of CS or BS were fixed, CaCO₃ content, deviatoric stress, shear strength, cohesive force, and internal friction angle all showed a trend of first increasing and then decreasing with the increase in CS or BS concentration. The optimal combination of CS and BS concentration was 1.5 mol/L and OD₆₀₀ = 1.8. Thermochemical analyses showed an improved thermal stability of the reinforcing cohesive soil, with the lowest mass loss (32%) and the highest pyrolysis temperature (812 °C) of the samples at the optimal combination of BS and CS concentration. This study is expected to improve the understanding of the MICP reinforcement process and contribute to the optimal design of future biologically mediated soil amendments, promoting bioremediation.

Fractional vegetation cover (FVC) is a core indicator for evaluating ecological environment quality, soil conservation capacity and riparian ecosystem sustainability. Using 30 m Landsat time-series data spanning 2011–2025, this study classified FVC into five grades. The Theil–Sen median slope estimation, Mann–Kendall significance test, univariate linear regression, and wavelet coherence method were applied to investigate spatiotemporal dynamics and time–frequency coupling responses of multi-grade FVC to the driving factors of sustainable ecological governance. Results demonstrated that 63.4% of the area exhibited stable vegetation coverage, 28.3% presented significant improvement, and only 8.3% suffered degradation. Spatially, high- and medium-high-FVC areas were concentrated in the riparian ecological zone, artificial forests, and agricultural land; while low-FVC areas were concentrated in urban and industrial zones. Linear regression evidenced a statistically significant increasing trend for high-FVC areas (R² = 0.304, p = 0.033), whereas low-FVC areas presented non-significant long-term monotonic variation (R² = 0.079, p = 0.309). The results of wavelet transform coherence tests revealed that high-FVC areas maintained persistent in-phase synchronous coupling with agricultural irrigation volume on a two-year cycle, and exhibited anti-phase lagged oscillations against total water resource volume. Low-FVC areas were sensitive to agricultural irrigation volume, and the influences of total water resources volume and annual precipitation on FVC were time-varying rather than persistent. This study provides a quantitative theoretical and methodological support for sustainable ecological spatial management and riverbank buffer zone construction in similar urbanized river basins.