Accurate mapping of tea plantations in fragmented, mountainous landscapes faces challenges from spectral confusion, cloud-induced data gaps, and limited model transferability. To address these issues, this study proposes a data-driven approach leveraging 64-dimensional Google AlphaEarth Foundations (AEF) satellite embeddings as core predictive features, integrated with Sentinel-2 spectral, textural, and topographic variables. Prior to feature optimization, comparative experiments confirmed that Random Forest outperformed Gradient Boosting Trees, Classification and Regression Trees, and Support Vector Machines in stability and accuracy, serving as the core classifier. Leveraging a robust sampling strategy, this study evaluated 12 classification scenarios. Results showed that the AEF-augmented scenario achieved the best performance in Rizhao (Overall Accuracy 92.69%, Kappa 0.90), with a high Producer’s Accuracy of 97.47% that effectively minimized omission errors. SHapley Additive exPlanations (SHAP) analysis revealed the model’s physically interpretable logic: utilizing embeddings as “exclusion filters” to separate tea from non-target classes by encoding latent phenological patterns, while relying on original spectral bands to capture canopy biological signals. Crucially, the model demonstrated exceptional generalizability when transferred to the unseen Qingdao region without retraining. This study validates AEF embeddings as a robust, scalable feature representation for regional crop monitoring in label-scarce and heterogeneous environments, offering a transferable data foundation for precise agricultural inventory and sustainable development planning.

Urban flooding has become a critical environmental challenge under global climate change and rapid urbanization. This study develops a multi-indicator hazard mechanism framework for flood hazard assessment in Rizhao, a coastal city in China, by integrating three fundamental hydrological processes: runoff generation, flow convergence, and drainage. Based on geospatial data—including DEM, road networks, land cover, and soil characteristics—six key indicators were evaluated using the TOPSIS method: runoff curve number, impervious surface percentage, topographic wetness index, time of concentration, pipeline density, and distance to rivers. The results show that extreme-hazard zones, covering 6.41% of the central urban area, are primarily clustered in northern sectors, where flood susceptibility is driven by the synergistic effects of high imperviousness, short concentration time, and inadequate drainage infrastructure. Independent validation using historical flood records confirmed the model’s reliability, with 83.72% of documented waterlogging points located in predicted high-hazard zones and an AUC value of 0.737 indicating good discriminatory performance. Based on spatial hazard patterns and causal mechanisms, an integrated mitigation strategy system of “source reduction, process regulation, and terminal enhancement” is proposed. This strategy provides practical guidance for pipeline rehabilitation and sponge city implementation in Rizhao’s resilience planning, while the developed hazard mechanism framework of “runoff–convergence–drainage” provides a transferable methodology for flood hazard assessment in large-scale urban environments.