Warming-induced drought stress and El Nino-associated summer precipitation failure are responsible for increased forest fire intensities of tropical and temperate forests in Asia and Australia. However, both effects are unclear for boreal forests, the largest biome and carbon stock over land. Here, we combined fire frequency, burned area, and climate data in the Altai boreal forests, the southmost extension of Siberia’s boreal forest into China, and explored their link with El Nino–Southern Oscillation (ENSO). Surprisingly, both summer drought severity and fire occurrence showed significant (p < 0.05) correlation with La Nina events of the previous year and therefore provide an important reference for forest fire prediction and prevention in Altai. Despite a significant warming trend, the increased moisture over Altai has largely offset the effect of warming-induced drought stress and led to an insignificant fire frequency trend in the last decades, resulting in largely reduced burned area since the 1980s. The reduced burned area can also be attributed to fire suppression efforts and greatly increased investment in fire prevention since 1987.

The rubber tree (Hevea brasiliensis) is an important tree species for the production of natural latex, which is an essential raw material for varieties of industrial and non-industrial products. Rapid and accurate identification of the number of rubber trees not only plays an important role in predicting biomass and yield but also is beneficial to estimating carbon sinks and promoting the sustainable development of rubber plantations. However, the existing recognition methods based on canopy characteristic segmentation are not suitable for detecting individual rubber trees due to their high canopy coverage and similar crown structure. Fortunately, rubber trees have a defoliation period of about 40 days, which makes their trunks clearly visible in high-resolution RGB images. Therefore, this study employed an unmanned aerial vehicle (UAV) equipped with an RGB camera to acquire high-resolution images of rubber plantations from three observation angles (−90°, −60°, 45°) and two flight directions (SN: perpendicular to the rubber planting row, and WE: parallel to rubber planting rows) during the deciduous period. Four convolutional neural networks (multi-scale attention network, MAnet; Unet++; Unet; pyramid scene parsing network, PSPnet) were utilized to explore observation angles and directions beneficial for rubber tree trunk identification and counting. The results indicate that Unet++ achieved the best recognition accuracy (precision = 0.979, recall = 0.919, F-measure = 94.7%) with an observation angle of −60° and flight mode of SN among the four deep learning algorithms. This research provides a new idea for tree trunk identification by multi-angle observation of forests in specific phenological periods.

Prescribed burning is a widely used fuel management employed technique to mitigate the risk of forest fires. The Pinus yunnanensis Franch. forest, which is frequently prone to forest fires in southwestern China, serves as a prime example for investigating the effects of prescribed burning on the flammability of surface dead fuel. This research aims to establish a scientific foundation for managing dead fuel in forests, as well as fire prevention and control strategies. Field data was collected from P. yunnanensis forests located in central Yunnan Province in 2021 and 2022. The study implemented a randomized complete block design with two blocks and three treatments: an unburned control (UB), one year after the prescribed burning (PB1a), and three years after the prescribed burning (PB3a). These treatments were evaluated based on three indices: surface dead-bed structure, physicochemical properties, and potential fire behavior parameters. To analyze the stand characteristics of the sample plots, a paired t-test was conducted. The results indicated no significant differences in the stand characteristics of P. yunnanensis following prescribed burning (p > 0.05). Prescribed burning led to a significant decrease in the average surface dead fuel load from 10.24 t/ha to 3.70 t/ha, representing a reduction of 63.87%. Additionally, the average fire−line intensity decreased from 454 kw/m to 190 kw/m, indicating a decrease of 58.15%. Despite prescribed burning, there were no significant changes observed in the physical and chemical properties of dead fuels (p > 0.05). However, the bed structure of dead fuels and fire behavior parameters exhibited a significant reduction compared with the control sample site. The findings of this study provide essential theoretical support for the scientific implementation of prescribed burning programs and the accurate evaluation of ecological and environmental effects post burning.

Introduction: Positioning studies on prescribed burning in Pinus yunnanensis forests have been conducted for several years, focusing on the effects of fire on the composition and structure, growth, regeneration, relative bark thickness, and bark density of understory oak species in Pinus yunnanensis forests.Methods: The study was conducted on Zhaobi Mountain, Yi-Dai Autonomous County of Xinping, Yuxi City, Yunnan Province. In the prescribed burn after restoration of full 1 year of the area and did not implement the prescribed burn area were set up 10 m × 10 m sample plots 30 pairs of comparisons, and all the oak trees in the sample plots were recorded, each sample plot in the four apexes and the middle were set up five 2 m × 2 m small sample squares, the shrubs in the small sample squares for each plant survey, comparison, statistics and analysis of all data.Results: The study results showed that (1) prescribed burning significantly affected the species composition of the understorey of Pinus yunnanensis forests. In both tree and shrub layers, the important values of Quercus aliena, Quercus serrata, Quercus fabri, and Quercus variabilis were significantly reduced in the burned areas. In contrast, the important values of Quercus acutissima increased somewhat. (2) The under crown height of oak trees in the burned areas was significantly lower than in the burned areas, but the height of oak trees in the burned areas was not significantly different from that in the burned areas. In the shrub layer, the height and cover of oak plants in the prescribed burning areas were significantly lower than in the unprescribed burned areas, effectively reducing the vertical continuity of the forest surface combustible material and reducing the possibility of fire converting from surface to canopy fire along the “ladder fuel.” (3) The regeneration of oak plants in the burned area is mainly by sprout tillers, and very few young sprouts are regenerated by seed germination. Renewed young sprouts are difficult to survive the prescribed burn areas the following year due to their lack of fire tolerance. (4) The relative bark thickness and density of oak plants in prescribed burn areas were significantly higher than those in unprescribed burn areas due to the fire tolerance exhibited by oak plants in long-term prescribed burns.Discussion: Prescribed burning has profoundly altered the structural composition and growth of oak plants in the understory of Pinus yunnanensis forests, and oak plants have shown significant fire-adapted traits to resist fire under long-term fire disturbance. The study can provide a scientific basis for prescribed burning, forest fuels, and forest fire management.

Forest fire susceptibility prediction is essential for effective management, yet considerable uncertainty persists under future climate change, especially in climate-sensitive plateau regions. This study integrated MODIS fire data with climatic, topographic, vegetation, and anthropogenic variables to construct an Extreme Gradient Boosting (XGBoost) model for the Yunnan Plateau, a region highly prone to forest fires. Compared with Support Vector Machine and Random Forest models, XGBoost showed superior ability to capture nonlinear relationships and delivered the best performance, achieving an AUC of 0.907 and an overall accuracy of 0.831. The trained model was applied to climate projections under SSP1-2.6, SSP2-4.5, and SSP5-8.5 to assess future fire susceptibility. Results indicated that high-susceptibility periods primarily occur in winter and spring, driven by minimum temperature, average temperature, and precipitation. High-susceptibility areas are concentrated in dry-hot valleys and mountain basins with elevated temperatures and dense human activity. Under future climate scenarios, both the probability and spatial extent of forest fires are projected to increase, with a marked expansion after 2050, especially under SSP5-8.5. Although the XGBoost model demonstrates strong generalizability for plateau regions, uncertainties remain due to static vegetation, coarse anthropogenic data, and differences among climate models.

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.