Chinese forest ecosystems are key carbon sinks that significantly contribute to lowering carbon emissions. Accurate Net Ecosystem Productivity (NEP) estimations are essential for evaluating their carbon sequestration capabilities and overall health. This study employed the Physiological Principles Predicting Growth-Satellites (3-PGS) and soil heterotrophic respiration models to simulate China’s forest carbon sinks and sources distribution from 2013 to 2023. Then, climatic factors influencing NEP changes were examined through the application of a geographical detector model. The net carbon sequestered was 1.71 ± 0.09 PgC with an annual average of 0.156 ± 0.0071 PgC, signifying a substantial carbon sink in China’s forest. The annual NEP was highest in evergreen broadleaf forests (352.12 gC m⁻²) and lowest in deciduous needleleaf forests (148.31 gC m⁻²). NEP in China’s forests increased by a rate of 1.67 gC m⁻² annually, with most regions exhibiting a 275.32 gC m⁻² annual carbon sink. The geographical detector model analysis showed that solar radiation, precipitation, and vapor pressure deficit were the main drivers of NEP change, while temperature and frost days had a secondary influence. Furthermore, the interaction between solar radiation and temperature variables showed the greatest impact. This study can enhance the understanding of carbon sink and source distribution in China, serve as a reference for regional carbon cycle research, and provide key insights for policymakers in developing effective climate strategies.

Wildfires threaten natural ecosystems and human livelihoods in the Tanzanian Miombo woodlands. This study presents the first locally calibrated, high-resolution wildfire susceptibility map for the Iringa region, developed using a robust machine learning framework. Multi-decadal remote sensing data (MODIS fire occurrences, 2001–2024) were integrated with climatic, topographic, vegetation, and anthropogenic variables to train four classifiers: Random Forest, XGBoost, support vector machine with RBF kernel, and Logistic Regression. A balanced dataset of 9096 fire points and an equal number of randomly sampled non-fire points was used. The data were split into 70% for training and 30% for testing. Model performance was evaluated using accuracy, area under the ROC curve (AUC), accuracy, precision, and F1-score. Random Forest achieved the highest overall performance (AUC = 0.845, accuracy = 0.759, precision = 0.789 and F1 = 0.771), followed by XGBoost (AUC = 0.828, accuracy = 0.736, precision = 0.700 and F1 = 0.757), SVM (AUC = 0.755, accuracy = 0.679, precision = 0.648 and F1 = 0.709), and Logistic Regression (AUC = 0.740, accuracy = 0.661, precision = 0.631 and F1 = 0.696). Feature importance analysis identified altitude as the most influential variable, followed by wind speed, distance to road, and NDVI. Kernel Density Estimation revealed spatially distinct fire clusters concentrated in central and southern hotspots. Temporal analysis showed that 94% of fires occur during the dry season (June–November), peaking sharply in October. These findings provide an evidence-based framework for fire prevention and sustainable management of Iringa’s Miombo woodlands.

Forest cover monitoring, mapping, and change detection are fundamental to sustainable forest management and conservation. Although advances in geospatial and remote-sensing technologies have substantially enhanced the capacity to analyze large, long-term datasets, assessments of forest-cover change in Kenya have largely remained localized. Consequently, integrated national-scale, multi-decadal analyses that link land-cover transitions, vegetation condition, and interacting drivers remain scarce. This study employed time-series MODIS datasets within Google Earth Engine to quantify spatiotemporal forest-cover dynamics across Kenya from 2001 to 2023 and to characterize vegetation condition using NDVI and Leaf Area Index (LAI). Driver attribution was conducted using GeoDetector to evaluate the explanatory power (q-statistics) of anthropogenic and natural factors and to assess their nonlinear interactions. Results indicate a net forest-cover gain of approximately 6977.8 km² over the study period; however, transition analyses and vegetation-index trajectories reveal widespread forest degradation and increasing structural openness of forest landscapes across several regions. GeoDetector results show consistently stronger explanatory power of anthropogenic drivers relative to climatic variability across multiple periods. These findings reveal that net forest-area gains can coincide with ongoing degradation, underscoring the necessity for sustained, spatially targeted monitoring and integrated management.