Ecosystem service value (ESV) is a crucial indicator for evaluating ecosystem health, and identifying its spatial driving factors will help to provide scientific decision support for ecological protection and restoration. This study took the Liuxi River Basin in China as the research object and used the value equivalent method to estimate regional ESV. In the process of using the Geodetector model (GDM), the study area was spatially stratified by using the local bivariate spatial correlation pattern to mine the potential driving factors of ESV. The results show that: (1) From 2005 to 2018, the total value of ecosystem services in the Liuxi River Basin showed a fluctuating and increasing trend. ESV had high-value aggregation in the northeastern mountainous areas with high green space coverage and high river distance accessibility and low-value aggregation in the central and southwestern urban areas with frequent human activities. Its spatial heterogeneity and aggregation patterns were of statistical significance. (2) The spatial distribution characteristics of ESV were affected by various driving factors to varying degrees. The order of their degree of influence on ESV was per capita green area > slope > the proportion of urban and rural human settlements > river distance accessibility > population. (3) Compared to the previous study, the stratification method employing the local bivariate spatial correlation pattern more fully considers spatial autocorrelation and spatial heterogeneity. It effectively captured the spatial explanatory power of driving factors. This study can provide new ideas for capturing the driving mechanisms of ESV and insights into the sustainable development of the ecological environment in other regions with similar characteristics worldwide.

Accurate mapping of built-up land is essential for urbanization monitoring and ecosystem research. At present, remote sensing is one of the primary means used for real-time and accurate surveying and mapping of built-up land, due to the long time series and multi-information advantages of existing remote sensing images and the ability to obtain highly precise year-by-year built-up land maps. In this study, we obtained feature-enhanced data regarding built-up land from Landsat images and phenology-based algorithms and proposed a method that combines the use of the Google Earth Engine (GEE) and deep learning approaches. The Res-UNet++ structural model was improved for built-up land mapping in Guangdong from 1991 to 2020. Experiments show that overall accuracy of built-up land map in the study area in 2020 was 0.99, the kappa coefficient was 0.96, user accuracy of built-up land was 0.98, and producer accuracy was 0.901. The trained model can be applied to other years with good results. The overall accuracy (OA) of the assessment results every five years was above 0.97, and the kappa coefficient was above 0.90. From 1991 to 2020, built-up land in Guangdong has expanded significantly, the area of built-up land has increased by 71%, and the proportion of built-up land has increased by 3.91%. Our findings indicate that the combined approach of GEE and deep learning algorithms can be developed into a large-scale, long time-series of remote sensing classification techniques framework that can be useful for future land-use mapping research.

The frequent occurrence of urban floods (UFs) poses significant threats to public safety and the national economy. Accurate estimation of urban flood susceptibility (UFS) and the identification of potential hotspots are critical for effective UF management. However, existing UFS studies often fall short due to a limited understanding of UFs’ nature, frequently relying on disaster factors analogous to those used for natural floods while neglecting key urban characteristics, limiting the accuracy of UFS estimates. To address these challenges, we propose a novel framework for UFS assessment. Unlike those studies that focus primarily on topographic and surface characteristics, our approach integrates urban-specific factors that capture the distinctive attributes of the urban environment, including Urban Heat Island Intensity, Urban Rain Island Intensity, Urban Resilience Index, and Impervious Surface Percentage. Guangzhou was selected as the study area, where machine learning methods were employed to calculate UFS, and Shapley Additive Explanation was utilized to quantify the contributions of employed factors. We evaluated the significance of urban factors from three perspectives: classifier performance, map accuracy, and factor importance. The results indicate that (1) urban factors hold significantly greater importance compared to other factors, and (2) the incorporation of urban factors markedly enhances both the performance of the trained classifier and the accuracy of the UFS map. These findings underscore the value of integrating urban factors into UFS assessments, thereby contributing to more precise UF management and supporting sustainable urban development.

Forest fire events in coastal eastern Guangdong are strongly associated with human activities and often occur as small-scale, short-lived early-stage fire events. Given uneven thermal monitoring infrastructure, a key challenge is leveraging data-rich local areas to identify high-susceptibility areas in data-sparse regions. Focusing on Shanwei City, this study used fire data from thermal imaging in the Shenzhen–Shanwei Special Cooperation Zone. By integrating biophysical covariates and anthropogenic factors, we developed a Randomized Iterative Sampling–Similarity Assessment–Model Selection (RISAMS) framework to optimize negative-sample selection and identify a robust classifier. Shapley Additive exPlanations (SHAP) was implemented in Python 3.13.5 using the SHAP package (version 0.48.0) and was used to interpret drivers of susceptibility. In the optimal iteration, mean covariate similarity between training samples and the Shanwei macro-environmental background reached 0.94, with a minimum of 0.72. The optimal Random Forest classifier achieved an AUC of 0.92 on the training set and 0.91 on the test set, with Cohen’s Kappa of 0.78 for both. High-susceptibility zones clustered along forest edges and wildland–urban interfaces in northern and central Shanwei. SHAP identified road density, elevation, and slope as the primary drivers, with a cumulative relative importance of 61.87%. This framework supports early-stage forest fire susceptibility mapping under uneven monitoring resources and improves early warning of early-stage fires in human-dominated humid coastal forests.