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.

Preseason temperature has always been considered the most critical factor influencing vegetation phenology in the northern hemisphere. While numerous studies have examined the impact of daytime and nighttime warming on vegetation phenology in this region, the specific influence of day and night warming on deciduous broad-leaved forests (DBFs) in Northern China, where significant temperature variations occur between day and night, remains unclear. Furthermore, the sensitivity of daytime and nighttime warming during different preseason periods to phenology has not been quantitatively understood. We analyzed GIMMS3g NDVI data from 1985 to 2015 and employed a double logistic regression model to determine the phenological start of the season (SOS) for DBF in Northern China. To control for monthly precipitation effects, we conducted partial correlation analysis between monthly mean maximum daytime temperature (T_{day_max}), monthly mean minimum nighttime temperature (T_{night_min}), diurnal temperature variation (DTR), and SOS. Our findings over the past 31 years indicate that 75.98% of the area exhibited an advanced trend, with an overall advance of 1.7 days per decade. Interestingly, regardless of T_{day_max}, T_{night_min}, or DTR, most areas had a preseason length of 1 month, accounting for 50.26%, 34.45%, and 44.39%, respectively. Furthermore, approximately 50.68% of the area exhibited a significant negative correlation between preseason temperature and SOS for T_{day_max}, 34.02% for T_{night_min}, and 35.80% for DTR. It can be found that the response of the SOS advance to T_{day_max} in DBFs in Northern China is more obvious than that to T_{night_min} and DTR. Our study revealed that the difference in day and night temperature warming on DBFs in Northern China is not pronounced. Specifically, SOS advanced by 1.8 days, 1.98 days, and 1.95 days for every 1 °C increase in T_{day_max}, T_{night_min}, and DTR, respectively. However, it is important to note that the distribution of advanced days resulting from the warming of these three preseason temperature indicators exhibited spatial heterogeneity. Although many studies have already established the influence of various meteorological indicators on spring phenology, determining which meteorological indicators should be employed to quantify their impact on phenology in different regions and vegetation types remains a subject for further exploration and investigation in the future.

Vegetation plays a dual role in the Earth’s climate system: it removes atmospheric CO₂ through photosynthesis while emitting biogenic volatile organic compounds (BVOCs), which can weaken the net carbon sink and contribute to air pollution. To assess the long-term interplay between carbon uptake and BVOC emissions, and to clarify how vegetation characteristics and climate regulate this relationship, we developed a Biogenic Carbon Efficiency Index (BCEI). The BCEI integrates BVOC emissions with gross primary productivity (GPP) to quantify their spatial ratio, thereby capturing the concurrent “source” and “sink” attributes of vegetation. We characterize the spatiotemporal heterogeneity of the BCEI across China and identify its dominant environmental drivers. The BCEI decreases from southeast to northwest, and during 2001–2020 exhibited a declining trend over 78% of the country, with increases mainly in Southwest China and on the Shandong and Liaodong Peninsulas. Driver analyses indicate that variables linked to hydrothermal conditions, including temperature, precipitation, evapotranspiration, and soil moisture, primarily control BCEI variability. Across most regions, the BCEI is negatively correlated with soil moisture and precipitation, positively correlated with evapotranspiration, and shows regionally varying associations with temperature. These findings deepen understanding of vegetation’s dual role as a source and sink and its driving mechanisms, providing a theoretical basis for optimizing regional vegetation management strategies.