Understanding post-fire vegetation recovery dynamics is crucial for damage assessment and recovery planning, yet spatiotemporal patterns in complex plateau environments remain poorly understood. This study addresses this gap by focusing on Yunnan Province, a mountainous plateau region with high fire incidence. We developed an innovative approach combining differenced Normalized Burn Ratio (dNBR) and visual interpretation on Google Earth Engine (GEE) to generate high-quality training samples from Landsat 5 TM/7 ETM+/8 OLI imagery. Four supervised machine learning algorithms were evaluated, with Random Forest (RF) demonstrating superior accuracy (OA = 0.90) for fire severity classification compared to Support Vector Machine (SVM) OA of 0.88, Classification and Regression Tree(CART) OA o f0.85, and Naive Bayes(NB) OA of 0.78. Using RF, we generated annual fire severity maps alongside the Land Surface Water Index (LSWI), Normalized Difference Vegetation Index (NDVI), and Normalized Burn Ratio (NBR) from 2005 to 2020. Key findings include the following: (1) fire severity classification outperformed traditional remote sensing indices in characterizing vegetation recovery; (2) distinct recovery trajectories emerged across severity levels, with moderate areas recovering in 7 years, severe areas transitioning within 2 years, and low severity areas peaking at 2 years post-fire; (3) southern mountainous regions exhibited 1–2 years faster recovery than northern areas. These insights advance understanding of post-fire ecosystem dynamics in complex terrains and support more effective recovery strategies.

This study aimed to accurately map burned forest areas and analyze the spatial distribution of forest fires under complex terrain conditions. This study integrates Landsat 8, Sentinel-2, and MODIS data to map burned forest areas in the complex terrain of western Yunnan. A machine learning workflow was developed on Google Earth Engine by combining Dynamic World land cover data with official fire records, utilizing a logistic regression-based feature selection strategy and an enhanced SNIC segmentation GEOBIA framework. The performance of four classifiers (RF, SVM, KNN, CART) in burn detection was evaluated through a comparative analysis of their spectral–spatial discrimination capabilities. The results indicated that the RF classifier achieved the highest performance, with an overall accuracy of 96.32% and a Kappa coefficient of 0.951. Spatial analysis further revealed that regions at medium altitudes (800–1600 m) and moderate slopes (15–25°) are more prone to forest fires. This study demonstrates a robust approach for generating accurate large-scale forest fire maps and provides valuable insights for effective fire management in complex terrain areas.

Olive (Olea europaea L.) is an important woody oil crop worldwide, and accurate estimation of leaf chlorophyll content is critical for assessing nutritional status, photosynthetic capacity, and precision crop management. Unmanned aerial vehicle (UAV) remote sensing, with high spatiotemporal resolution, has increasingly been applied in crop growth monitoring. However, the small, thick, waxy leaves of olive, together with its complex canopy structure and dense arrangement, may reduce estimation accuracy. To identify sensitive features related to olive leaf chlorophyll and to evaluate the feasibility of UAV-based estimation methods for olive trees with complex canopy structures, UAV multispectral orthophotos were acquired, and leaf chlorophyll was measured using a SPAD (Soil Plant Analysis Development) meter to provide ground-truth data. A dataset including single-band reflectance, vegetation indices, and texture features was built, and sensitive variables were identified by Pearson correlation. Modeling was performed with linear regression (LR), Random Forest (RF), Extreme Gradient Boosting (XGBoost), Partial Least Squares Regression (PLSR), and Support Vector Machine (SVM). Results showed that two spectral bands (green and red), one vegetation index (TCARI/OSAVI), and twelve texture features correlated strongly with SPAD values. Among the machine learning models, XGBoost achieved the highest accuracy, demonstrating the effectiveness of integrating multi-feature UAV data for complex olive canopies. This study demonstrates that combining reflectance, vegetation indices, and texture features within the XGBoost model enables reliable chlorophyll estimation for olive canopies, highlighting the potential of UAV-based multispectral approaches for precision monitoring and providing a foundation for applications in other woody crops with complex canopy structures.

Carbon stock (CS) is an important indicator of the structure and function of forest ecosystems, and plays an important role in mitigating climate change, maintaining ecological system balance, promoting carbon trading, and other socioeconomic and ecological values. Olea europaea L. is a species of high economic and ecological value, and its excellent nutritional composition, strong drought tolerance, sustainable production characteristics, and promotion of agrodiversity make it important in guaranteeing food security. Accurately estimating the CS of Olea europaea L. offers a reliable reference for its artificial breeding and yield prediction. Firstly, an independent estimation model of Olea europaea L. CS was constructed, while a compatibility model of Olea europaea L. unitary and binary CS was constructed using nonlinear metric error. Secondly, in the CS compatibility model system, the total CS model of Olea europaea L. was constructed by the Logarithmic Nonlinear Seemingly Unrelated Regression (LNSUR) method with D and D²H as independent variables. The results show: (1) The independent model of Aboveground CS (AGCS) was C = 0.0014D^1.92876H^0.67174 (R² = 0.909), and the independent model of Belowground CS (BGCS) was C = 0.00723D^1.23578H^0.48553 (R² = 0.686). The AGCS compatibility model effectively addresses the issue of component sums not equaling the total, while maintaining a low RMSE (1.918); (2) The LNSUR model improved the accuracy of the BGCS model more significantly (R² = 0.787), and the estimated total CS also had a smaller RMSE (0.241~0.418); (3) Whole-plant CS of Olea europaea L. in 15 sample plots was estimated using the CS independent model and the LNSUR model with an R² of 0.964. This study is the first attempt to construct a CS estimation model for Olea europaea L., which provides a scientific and technological basis for the monitoring of its economic and ecological value indicators, such as yield and carbon sink capacity.