The low-temperature bonding performance at the asphaltmastic -aggregate interface plays a crucial role in determining the low-temperature crack resistance and water stability of asphalt mixtures in cold regions. This study employs a pull-off test coupled with ImageJ software for detailed image analysis of the failure surfaces. It focuses on two key metrics： bonding strength and bonding failure ratio， to evaluate the bond performance of asphalt， asphalt mortar， and asphalt concrete with limestone aggregates under varying low temperatures （-10， -20， and -30 ℃） and in wet conditions. The results reveal that the asphalt-aggregate interface achieves the highest bonding strength at -20 ℃. However， the bonding failure ratio increases as temperatures decrease. Notably， the presence of water can lead to a 49% reduction in low-temperature bonding strength， resulting in a continuous detachment of the interface and an increase in the bonding failure ratio by 44%. Furthermore， an optimal powder-binder ratio can enhance the low-temperature bonding strength of asphalt mortar； specifically， a ratio of 1.2 yields a bonding strength that is 1.44 times greater than that of the asphalt matrix. Consistency in the low-temperature bonding strength results is further corroborated through bending creep stiffness tests. The incorporation of mineral powder and fine aggregates modifies the contact characteristics at the interface， leading to significant alterations in the failure location and bonding failure ratio. Failures in asphalt mortar typically occur within the mortar itself， while those in asphalt concrete predominantly occur at the interface， with the bonding failure ratios of asphalt concrete being substantially higher than those of asphalt mortar.

In view of the limitations of existing indicators， in order to better evaluate the high and low temperature performance of warm mix asphalt binder， different amounts of Sasobit and Evotherm 3G warm mix agent were selected to be mixed with 70^# matrix asphalt and SBS modified asphalt to prepare modified asphalt. Two parameters of complex shear modulus G^* and phase angle δ were obtained and analyzed by dynamic shear rheological （DSR） test. The rutting factor G^*/sinδ， improved rutting factor G^* /（sinδ）⁹ index and critical temperatureT_{G\mathrm{*}/\mathrm{s}\mathrm{i}\mathrm{n}\delta }， improved critical temperatureT_{G\mathrm{*}/{(\mathrm{s}\mathrm{i}\mathrm{n}\delta )}^{\mathrm{9}}} index were used to comprehensively evaluate and analyze the high temperature performance of modified asphalt. Two parameters of creep stiffness modulus S and creep rate m were obtained and analyzed by low temperature bending rheological （BBR） test. The k index was established and the creep compliance J（t） index was introduced to evaluate the low temperature performance of modified asphalt. The test results and data analysis showed that： G^* /（sinδ）⁹ was more accurate than G^*/sinδ in evaluating the high temperature performance of warm mix asphalt， and T_{G\mathrm{*}/{(\mathrm{s}\mathrm{i}\mathrm{n}\delta )}^{\mathrm{9}}} was suitable for the high temperature performance evaluation of SBS warm mix asphalt， and there was no obvious difference for 70^# matrix asphalt. The k index can distinguish the difference of low temperature performance between matrix asphalt and modified asphalt， and the J（t） index can well reflect the low temperature creep performance of warm mix asphalt. Finally， the weight analysis of high and low temperature indexes was carried out by analytic hierarchy process （AHP）， and the weight values of T_{G\mathrm{*}/{(\mathrm{s}\mathrm{i}\mathrm{n}\delta )}^{\mathrm{9}}} and J（t） were the largest. It was suggested that T_{G\mathrm{*}/{(\mathrm{s}\mathrm{i}\mathrm{n}\delta )}^{\mathrm{9}}} was used to evaluate the high temperature performance of warm mix asphalt， and J（t） was used to evaluate the low temperature performance of warm mix asphalt.

Due to the high intensity and frequency of rainfall throughout the year in tropical rainforest areas， heavy rainfall often leads to an increase in the self weight of the rock and soil mass， but significantly reduces its strength and stability， which is the main high risk source for tunnel construction in this area. Based on a tunnel project in the tropical rainforest region of Malaysia， field monitoring and numerical analysis were conducted on the deformation behavior of primary support clearance limit in the soil-rock transition section， investigating emergency measures for controlling primary support clearance limit and implementing local arch replacement construction. The research findings were as follows： （1） The primary causes of significant tunnel deformation in the tropical rainforest region were reduced rock and soil strength and bearing capacity resulting from frequent and intense rainfall infiltration. The variation of formation water content was mainly affected by rainfall time and surface fluctuation. （2） Measures such as radial grouting， counter-pressure backfilling， temporary transverse bracing， and temporary inverted arch construction effectively controlled the development of surrounding rock deformation. （3） A method for local arch replacement construction was used based on the longitudinal distribution pattern of deformation clearance limit in the primary support of the tunnel， resulting in improved efficiency and safety. On-site applications had demonstrated that the adopted encroachment limit control method had a good effect on controlling tunnel deformation， providing valuable construction references and technical guidance for tunnels in similar tropical rainforest areas with weak rock formations.

The tunnel structure undergoes gradual changes in stress and deformation characteristics over time due to various complex factors during construction and operation. To address this， a deep learning probabilistic model incorporating an attention mechanism is proposed to accurately predict and assess the safety status at critical adverse locations in the tunnel lining. Initially， the Spearman rank correlation coefficient is employed for data preprocessing to select soil pressure and concrete strain data， which are highly correlated with the most adverse locations in the lining structure， as input features. Subsequently， a multi-layer convolutional neural network （CNN） is designed for multi-source data feature extraction， and a feature-sharing layer is constructed to integrate data information from different locations. The extracted features are then fed into a long short-term memory （LSTM） network for time-series analysis and prediction， with an attention mechanism introduced to optimize feature weighting， thereby further improving prediction accuracy. Finally， a Gaussian probabilistic regression model is established to address the quantification and evaluation of uncertainty in safety factor calculations due to prediction errors in structural response. Based on data from a real tunnel engineering project， the response prediction results for adverse locations indicate that the model can comprehensively account for the spatio-temporal correlation of multi-source measurement data. The average prediction errors for concrete strain on the training， validation， and test sets are 0.89， 1.02， and 1.24 \mathrm{\mu }\mathrm{\epsilon }， respectively， with no overfitting observed， verifying the proposed method's strong generalization capability in handling complex nonlinear problems. Additionally， a Gaussian probabilistic interval prediction approach is adopted， and a 90% confidence interval estimate is performed using the predicted safety factors. The results show that the safety factors for three critical locations in the secondary lining fall within this confidence interval， further validating the reliability and practicality of the proposed model in assessing the safety of tunnel lining structures.

Sand aggregate is a non-renewable resource， but the contradiction between supply and demand between the shortage of sand and stone resources and the large-scale expansion of the construction field is escalating. Under the background of the dual carbon policy of various countries， the application of abundant aeolian sand and pebble resources in northwest China to pile aggregate is an important means to achieve green environmental protection， cost reduction and efficiency increase in the engineering field. Based on the new pebble-aeolian sand concrete mix ratio obtained in the previous test， the concrete model pile was poured， and the bearing performance of the new pebble-aeolian sand concrete pile combined with soil was studied by model test. Results showed that： the optimized ratio of pebble-eolian sand concrete can better meet the strength requirements of composite foundation. Increasing pile length and decreasing pile diameter and pile spacing were helpful to reduce the settlement of foundation. Under 400kPa loading， the pile length was doubled， and the total settlement of the foundation was reduced by 89%. Compared with the condition of decreasing pile diameter and pile spacing， the settlement decreased by 24% and 36.4%， respectively. For the pile body stress， increasing the pile length， pile diameter， reducing the pile distance， the stress showed a trend of increasing. For pile soil stress， increasing pile length， pile diameter and decreasing pile distance can reduce the stress level of foundation soil. For the pile-soil stress ratio， the longer the pile length， the greater the pile-soil stress ratio. The research results of this paper can provide data support for the popularization and application of pebbles and aeolian sand.

In order to improve the forest residue pulverizing efficiency， pulverizing uniformity and machine operation stability， the article designed a curved beveled blade pulverizing tool for forest residue pulverizing. Through the mechanical and kinetic analysis of the shear and impact effects on the forest residue pulverizing process， the main factors affecting the pulverizing performance of the forest residue and the fragmentation mode were clarified. Based on discrete element simulation， the effects of straight-edged and beveled-edged knives on the pulverizing performance of forest residues were compared， and the simulation results showed that，when pulverizing forest residues under the same conditions， compared with straight-edged knives， the performance indexes such as the number of particle bond keys， average particle motion speed， average pulverizing power， impact on the walls of the pulverizing chamber， and average energy of the pulverizing particles of the beveled-edged knife simulation were all improved， and the degree of pulverizing using beveled-edged knives simulation was high， with a uniform pulverizing granularity， and a smoother working process.

In response to the threat posed by pine wood nematode disease to pine and other tree species， as well as the problems of transportation difficulties， low efficiency， high transmission risks， and poor crushing results inherent in traditional methods of handling large-diameter infected wood， a large-diameter infected wood crushing system was developed to enable on-site processing. Accordingly， a tracked chassis was designed for this crushing equipment； infected wood is first pretreated using hydraulic splitting technology and then crushed by a disc chipper followed by a hammer mill. Multibody dynamics simulation software was employed to simulate the operating behavior of the tracked chassis in complex terrains， including a 20° slope， a vertical wall measuring 200 mm in height and 400 mm in width， and a ditch measuring 200 mm in depth and 400 mm in width. Key parameters such as translational acceleration， vertical acceleration， and pitch angle were analyzed during operation. Under these conditions， the tracked chassis exhibited strong stability and pass ability， with all key parameters remaining within acceptable ranges and no incidents of track derailment. This large-diameter infected wood crushing equipment enables localized crushing， providing insights and technical support for infected wood management and tracked chassis design.

Currently， although there are some bamboo slice defect detection schemes based on image processing techniques， these schemes detect fewer types of defects， are less practical， and are difficult to deploy on machines. For this reason， an improved defect detection model for bamboo slice is proposed. Therefore， we propose an improved model for bamboo slice defect detection. The model proposed in this paper is an improved Deformable-DETR model， which firstly replaces the original backbone extraction network ResNet with InternImage， which is stacked with DCNv3 convolution as the core. This network retains the a priori properties of the traditional CNN and captures the long-range dependencies， making the extracted feature spatial semantics richer. Then， after feature extraction， a new sampling module is added， which abstracts the image feature mapping into fine a fine foreground target feature vectors and a small number of coarse background context feature vectors， which can not only remove redundant backgroud feature information but also extract high-semantic foreground.Finally， a novel collaborative hybrid allocation training scheme is introduced， which supervises the training of multiple parallel auxiliary heads through one-to-many label allocation， to easily improve the encoder's learning capability in an end-to-end detector. In addition， data augmentation is used to extend the dataset and migration learning is used to enhance the detection of bamboo slice defects. The experimental results show that the method proposed in this paper improves the defective feature extraction and parsing ability of the model， achieves 85.7% of mAP50 on the test dataset， the inference time for a single image is 0.28 seconds， and the detection accuracy is better than other mainstream target detection models， which provide a new method for detecting defects in bamboo slices.

To enhance the qualified rate of bumper painting for forestry transportation vehicles， the coating quality data of bumpers from a certain company were selected for analysis. The coating quality data were analyzed by using a Pareto chart， and it was found that particles and orange peel were the main factors affecting the painting quality. The primary causes of particles and orange peel were analyzed from five aspects： personnel， machinery， materials， methods， and environment. Spearman's correlation coefficient was employed for feature extraction and importance analysis， revealing that factors such as the temperature and relative humidity in the paint spray booth（paint spray both temperature， relative humidity in paint spray booth）， robot spray flow rate， rotary cup rotation speed， spray distance， spray speed， paint brand， and robot all influenced the bumper's painting quality. By applying the classfication and regression tree （CART） algorithm， it was determined that spray flow rate， temperature and relative humidity in the paint spray booth， robots 2 and 4， and paint brand were the more critical factors affecting the bumper's painting quality. The results indicated that the analysis of painting quality data using the CART classification algorithm could effectively identify fault points， providing valuable insights for improving the quality of bumper painting.

In order to investigate the effect of drilling parameters （hole diameter， hole depth）， soil moisture content and soil hardness on the peak torque of the twist drill， single-factor and multi-factor tests were carried out by using the self-made drilling torque test bench. The test results showed that the peak drilling torque decreased linearly with soil moisture content， and increased linearly with soil hardness， drilling depth and drilling diameter. The optimal analysis of the regression model showed that withing the given factor level range， the minimum peak torque drilling parameter combination was 28% of soil moisture content， 10 mm of drilling diameter， 6 cm of drilling depth， and 0.3 N·m of peak torque. The maximum peak torque drilling parameter combination was 20% of soil moisture content， 14 mm of drilling diameter， 10 cm of drilling depth， and 0.8 N·m of peak torque. The deviation between the predicted value and the actual value of the peak drilling torque was less than 4%， and the drilling parameter results were reliable. The research results provide a theoretical basis for the dynamic selection of drill bit for lawn maintenance robots.

In order to solve the problem of inaccurate prediction of natural falling direction during chain-saw logging operations， which is prone to safety accidents and cause casualties and property losses， a method for determining the natural falling direction of trees based on smartphone image analysis is proposed. The target trees in the images are extracted using the Close-Form image matting algorithm based on K-means clustering algorithm improvement. The sub-pixel-centroid positioning method is used to determine the location of the tree center of gravity in a single image. The trees are fitted with three-angle and two-angle gravity center fitting according to the space vector composite and projection law， respectively， and calculate the natural falling direction. The experimental results show that there is no significant difference and high consistency between the results of three-angle， two-angle， and human experienced discrimination methods in judging the natural falling direction of trees （F=0.008， P=1.000>0.05； ICC is 0.990， P=0.000<0.05）. Because of the simplicity of the two-angle measurement， the two-angle discrimination method can be used to determine the natural falling direction of trees. Extended experiments show that the two-angle discrimination method has high accuracy （F=0.003， P=0.997>0.05）， and can provide a certain reference for accurately judging the natural falling direction of trees.

In order to explore the adaptation strategies of Taxus cuspidata to global warming， the seasonal dynamics and influencing factors of stable carbon isotope composition （δ¹³C） and water use efficiency （WUE） of Taxus cuspidata （seedling， sapling， mature tree） at different growth stages were analyzed based on stable isotope technique. The results showed that the δ¹³C values of Taxus cuspidata leaves at different growth stages ranged from -3.051% to -2.939%， with an average value of -2.981%±0.061%. WUE ranged from 58.96 to 71.68 μmol/mol， with an average of 66.87 μmol/mol±6.90 μmol/mol. The δ¹³C and WUE of Taxus cuspidata showed a decreasing trend of early growing season （June） >middle growing season （August） > late growing season （September）， and the mature tree>sapling>seedling， at different growth stages. There was a significant linear negative correlation between WUE and 10 cm soil water content in different growth stages of Taxus cuspidata （seedling， y=-0.82x+107.29， R²=0.80， P<0.01； sapling， y=-0.34x+84.17， R²=0.45， P<0.05； mature tree， y=-0.93x+101.32， R²=0.44， P<0.05）. Soil water content was the main controlling factor of WUE of Taxus cuspidata. Taxus cuspidata in different growth stagesselected different water use strategies according to individual water demand and the degree of influence by external water and heat factors.

To explore the identification method of tree species in planted forests based on the Bayesian optimization convolutional neural network （PCA-BO-CNN） algorithm model based on principal component analysis， to improve the accuracy and robustness of remote sensing technology in tree species identification in planted forests. In this study， the Sahanba Mechanical Forest Farm was selected as the study area. Sentinel-1 remote sensing data， Sentinel-2 remote sensing data， DEM data， and the forest resource category 2 survey data were combined with the PCA-BO-CNN algorithm model， and compared with other algorithm models， to improve the accuracy of tree species identification in planted forests. The results showed that： （1） Compared with the pre-PCA algorithm， the standard deviation of 39 features of PCA1-PCA39 of multi-source data features after PCA algorithm processing and the differentiation among features were significantly improved. Therefore， PCA was beneficial to improve the identification accuracy of the dominant species of Larix gmelinii var. principis-rupprechtii （Mayr） Pilger， Betula platyphylla Sukaczev， Pinus sylvestris var. mongholica Litv.， Quercus mongolica Fisch. ex Ledeb. and Picea asperata Mast. as well as non-forest land. （2） Before PCA algorithm， the overall accuracy （OA） and Kappa coefficient accuracy of the BO-random forest （BO-RF） algorithm model for the identification of dominant tree species and non-forest land were 81.87% and 0.754 5， respectively. After PCA algorithm processing， the accuracy of OA and Kappa coefficients of the PCA-BO-CNN algorithm model for the identification of dominant tree species and non-forest land was relatively improved， which were 83.10% and 0.770 3， respectively. （3） Compared with the BO-RF algorithm model before PCA algorithm processing， the overall accuracy of the PCA-BO-CNN algorithm model after PCA algorithm processing was relatively higher for the identification of the F₁， OA and Kappa coefficients of the main dominant tree species and non-forest land in Saihanba Forest Farm. Specifically， compared with the BO-RF algorithm model， the OA of the PCA-BO-CNN algorithm model increased by 1.24%. In addition， the OA of the PCA-BO-CNN algorithm model before PCA algorithm was improved by 3.71%. Compared with other algorithm models， the tree species identification method based on the PCA-BO-CNN algorithm model had strong accuracy and robustness， which can help us grasp the tree species distribution of the planted forests in Saihanba Forest Farm. Further， it provides an important methodological and theoretical basis for understanding forest carbon storage， forest response to climate change， formulating carbon emission reduction policies， and promoting forest sustainable development.

In order to reduce the browning phenomenon of Quercus mongolica during tissue culture， it was clarified whether the browning was related to the production of antioxidant enzyme system and phenolic substances. Mature zygotic embryos of Q. mongolica were used as materials， and the regeneration plant system of Q. mongolica was optimized by tissue culture technology to obtain regenerated plants. The effects of PVP on the physiological and biochemical indexes of zygotic embryo germination of Q. mongolica were investigated. The results indicated that PVP 40 had the best effect on inhibiting browning （40 represents the average molecular weight range of PVP）. The highest induction rate of adventitious buds was 71.17% when 0.2 g/L PVP 40 was added. The bud proliferation coefficient was 3.90， the rooting rate was 40.37%， and the transplanting survival rate was 73%. PVP reduced the activity of antioxidant enzymes， polyphenol oxidase and total phenolase activity of Quercus mongolica explants， and effectively inhibited browning. This study effectively reduced the browning effect of tissue culture of Q. mongolica and no longer aggravated the browning inhibition， optimized the organ regeneration plant system of Q. mongolica， and analyzed the internal physiological mechanism of PVP inhibition browning treatment on the germination of zygotic embryos of Q. mongolica， so as to provide reference for other species of Q. mongolica to inhibit browning.

Xinjiang is an important forest and fruit industry base in China， and the characteristic forest and fruit industry is an important part of regional economic development. In order to prevent fruit tree diseases from restricting the development of forest and fruit industry， a MobileNet-V2 NAM fruit tree leaf classification and disease identification model was designed in this study. It incorporated a lightweight normalization-based attention module to improve the model's sensitivity to feature information and make the model focus on salient features. At the same time， L1 regularization was added to the loss function to penalize the sparsity of the weights and suppress the non-significant weights. The experimental results showed that： in leaf classification， the model performed well in the classification results of self-built， Plant Village， and mixed datasets， with the accuracy rates reaching 97.05%， 98.73%， and 94.91%， respectively， and had good generalization ability. In disease identification， the MobileNet-V2 NAM model achieved a recognition accuracy of 94.55%， which was higher than the AlexNet， VGG16 classic CNN models， and the number of parameters of the model was only 3.56M. MobileNet-V2 NAM has good accuracy while maintaining a low amount of model parameters， provides technical support for embedding deep learning models into mobile devices.

Moss plays an important role in forest water conservation， nutrient fixation and water quality regulation. In order to reveal the influence of mosses on the water chemical characteristics of forest rainfall redistribution， this study selected Sphagnum and Pleurozium schreberi as research objects， and observed and studied the moss layer penetration rain and water chemical characteristics of the two mosses in the growing season from June to September of 2023. The results showed as follows： the maximum water holding capacity， maximum water holding rate and thickness of Sphagnum were significantly （P<0.01） higher than those of Pleurozium schreberi. There were differences in ion concentration between the two species of moss penetration rain. K⁺ showedthe most significant leaching effect of Sphagnum and F^- showedthe most significant retention effect. The K⁺ concentration increased by 241% after passing through the Sphagnum layer compared with the canopy penetration rain， while the F^- concentration decreased by 51.6%. The total metal ion concentration increased by 71.7% after passing through the Sphagnum layer， while the total concentration of nonmetallic ions decreased by 19.9%. \mathrm{P}{\mathrm{O}}_{\mathrm{4}}^{\mathrm{3}-} showed the most significant leaching effect of Pleurozium schreberi， and Ca²⁺ showed the most significant retention effect. \mathrm{P}{\mathrm{O}}_{\mathrm{4}}^{\mathrm{3}-} concentration increased by 147.7% after passing through the Pleurozium schreberi layer compared with the canopy penetration rain. The concentration of Ca²⁺ decreased by 5.2%. The total concentration of metal ions increased by 10.2% after passing through the Pleurozium schreberi layer， while the total concentration of non-metal ions decreased by 62.6%. （3） The leaching concentration of moss layer penetration rain was affected by pH value， temperature and canopy penetration rain ion concentration. Except Fe³⁺ and Cl^-， the other ions were all negatively correlated with the canopy penetration rain ion concentration. This study showed that the ion concentrations in the water of canopy penetration rain changed significantly after passing through the moss layer， which would in turn affect the chemical content of forest soil. The results of this study provide scientific support for the study of hydrology and hydrochemistry of the cold temperate forest in the Da Xing'anling Mountains.

As the main part of plant photosynthesis， chlorophyll plays an important role in monitoring vegetation growth and evaluating the capacity of carbon sequestration. Remote sensing technology， which is a high efficiency and low cost earth observation technology， can realize the estimation of chlorophyll content by leaf reflection spectrum characteristics. However， the accuracy of chlorophyll content estimation will be reduced due to the influence of leaf water content and leaf cell structure on leaf spectrum. Solar-induced chlorophyll fluorescence （SIF） is a technology of detecting the fluorescence information of chlorophyll excited fluorescence signal and the variation of fluorescence signal is directly related to chlorophyll content， which has great potential in chlorophyll content estimation. In this study， the fluorescence sensitive band of chlorophyll content was determined by sensitivity analysis using the fluorescence radiation transfer model SCOPE （Soil Canopy Observation， Photo-chemistry and Energy fluxes） and the chlorophyll content estimation model was established based on the fluorescence spectrum. Finally， the robustness of the model was verified by the measured data. The results showed that 700 nm and 730 nm were the highest and lowest sensitive bands of chlorophyll content， respectively. The band of 760 nm was the highest correlated band of chlorophyll content. The chlorophyll content estimation model were established based on the indices of fluorescence ratio of these three bands. SIF₇₆₀/SIF₇₀₀ had the best modeling accuracy with the R² of 0.998 1 and root mean square error （RMSE） of 0.043 5 μg/cm². The accuracy of SIF₇₀₀/SIF₇₃₀ and chlorophyll content （Cab） was the lowest in three models with R² and RMSE 0.904 8 and 0.088 6 μg/cm²， respectively. Independent samples of measured data were used to verify the above three estimated methods. SIF₇₆₀/SIF₇₃₀ had the best estimation results， with RMSE of 0.210 8 μg/cm²， followed by SIF₇₀₀/SIF₇₃₀， with RMSE of 0.345 4 μg/cm². But estimated model by using SIF₇₀₀/SIF₇₃₀ showed an overall overestimating results. The estimated results of SIF₇₆₀/SIF₇₀₀ showed a different results compared with measured data with RMSE of 0.743 5 μg/cm². In summary， the ratio vegetation index calculated by SIF₇₆₀/SIF₇₃₀ showed higher accuracy of modeling and excellent robustness of chlorophyll content estimation. Relevant studies provide technical references for the estimation of leaf biochemical parameters by chlorophyll fluorescence remote sensing technology.

Exploring the combined effects of forest management practices and environmental factors on the growth of Robinia pseudoacacia provides essential scientific evidence for afforestation and ecological restoration in the Yellow River Basin and other similar ecological environments. Focusing on Robinia pseudoacacia forests on the Loess Plateau， we measured topography， growth indicators of Robinia pseudoacacia， and soil physical and chemical properties， employing linear fitting， redundancy analysis， and Pearson correlation analysis to assess the impact of various factors on the growth of Robinia pseudoacacia. The results indicated： 1） Robinia pseudoacacia diameter at breast height （DBH） showed a highly significant negative correlation with afforestation density （P<0.01）， tree height had a significant negative correlation with afforestation density （P<0.05）， and canopy width was significantly negatively correlated with both afforestation density and slope （P<0.05）. 2） Tree height and canopy width of Robinia pseudoacacia showed a highly significant positive correlation with soil total phosphorus content （P<0.01）， while DBH and canopy width were significantly positively correlated with soil ammonium nitrogen content （P<0.05）. 3） At higher altitudes， the soil structure became more loose， but the nutrient content was relatively lower； when the slope was steeper， soil nutrient loss was severe， adversely affecting the growth of Robinia pseudoacacia. 4） Soil capillary porosity was the most important factor influencing the growth of Robinia pseudoacacia. Therefore， during afforestation and management， it is crucial to reasonably determine afforestation density and enhance soil fertility according to the actual topography and soil conditions to promote the growth and development of Robinia pseudoacacia.

The research aims to investigate the variations in negative air ions and air particulate matter in forest park， utilize the ecological functions of forests for the development of healthcare and wellness industries， and provide scientific perspectives for local economic development. This study selected Larix gmelinii forest， Betula platyphylla forest， Populus davidiana forest and Pinus sylvestris var. mongolica forest in Arctic Village National Forest Park as the research objects， and compared with the living area of Arctic Village. From May to October 2021， negative air ions， PM₁₀ and PM_2.5 and their influencing factors were observed on sunny and rainy days by using a portable air anion monitor. The results were as follows： 1） On sunny days， the daily dynamics of negative air ion concentrations in four forest types from large to small are manifested as leaf spreading period （LSP）， leaf opening period （LOP）， leaf falling period （LFP）. showing a unimodal curve trend， and the peak value appeared between 12：00 and 14：00， and the maximum can reach 4 207 ions/cm³. From large to small， they were Larix gmelinii forest， Betula platyphylla forest， Populus davidiana forest， Pinus sylvestris var. mongolica forest. In rainy days， the negative air ions concentration of the four forest types fluctuated， up to 6 000 ions/cm³， which was significantly higher than that of the living area （P<0.05）. On sunny days， the daily dynamic variation trend of PM₁₀ and PM_2.5 concentration of the four forest types was a bimodal curve， peaking in the morning and evening and reaching a trough at noon， but the fluctuation of the concentration on rainy days was small， which was far lower than that on sunny days （P<0.05）. 2） The monthly variation of the negative air ions concentration of the four forest types in May to October （sunny days） showed a unimodal curve trend， and the peak appeared in July， the maximum can reach 4 500 ions/cm^3， which was significantly higher than that of the living areas （P<0.05）. The air particulate matter concentration showed a trend of first decreasing and then increasing. 3） Negative air ions， PM₁₀ and PM_2.5 concentrations of the four forest types were significantly correlated with air temperature， air humidity and average wind speed. The negative air ion concentration in the forest ecosystem of the cold temperate forest park is between 2 000 and 6 000 ions/cm³， and the air particles are below 40 μg/m³， which is good air quality and conducive to people's physical and mental cultivation.

To optimize the preparation of amine-modified bamboo powder （AMBP）， the effects of alkaline pretreatment concentration， epoxidation time and temperature， amination reagent and modification temperature were investigated， and the adsorbent structure was further characterized using scanning electron microscopy （SEM）， Fourier transform infrared spectroscopy （FTIR）， X-ray photoelectron spectroscopy （XPS）， etc. The results showed that the alkaline pretreatment concentration， epoxidation temperature and time， amination temperature and modification reagents had significant effects on the adsorbent preparation. The optimal preparation conditions were as follows： 3.0 g of raw bamboo powder was treated with 2 mol/L NaOH alkaline， and then the epoxidation reaction was carried out in the mixture of 100 mL of 8% NaOH and 15 mL of epichlorohydrin at 40 ℃ for 8 hours； the final amination was conducted in the solution of 5 mL of TEPA， 1 g of Na₂CO₃ and 100 mL of water at 40 ℃ for 2 hours. The bamboo fiber structure of the original bamboo powder was well preserved during the modification process. After amine-modification， the N element mass fraction significantly increased， and the new functional groups such as —NH₂ and —NH— were introduced. The adsorbent AMBP-TEPA had a good adsorption performance on both free and complexed Cu （Ⅱ） in water. This study provides a new approach and technical support for the resource utilization of bamboo wood.

Wood bio-painting refers to using fungi and bacteria to change the color of wood， also called spalted wood. This article reviews its history， current status and challenges， and proposes prospects. The application of spalted wood can be traced back to Italy over 500 years ago， when it was widely used in inlay decoration techinques. Currently， research on spalted wood focuses on the study and utilization of zone line wood and pigments produced by soft rot fungi. The renewable and durable nature of microbial pigments offers a new approach to enhancing the comprehensive utilization rate and added value of plantation timber， and may help reduce the use of synthetic dyes. However， research of heartwood type（bacterial type） spalted wood poses greater challenges， requiring more scientific exploration and technological breakthroughs. When cultivating spalted wood through fungal inoculation， there are difficulties such as fungal strains preservation and decline， process pollution control， and cultivation condition. Additionally， pigment extraction and mordanting require attention to issues such as solvent toxicity and color fastness.

In tunnel engineering， ensuring the safety of the lining structure is crucial， with internal force analysis and safety assessment being key components. This study analyzed the structural internal forces through numerical simulations and on-site monitoring measurements， calculating the safety factors at critical locations. Given the existence of modeling and measurement errors， this paper introduced a probabilistic model to quantify the uncertainty in estimating safety factors， and normalizes the reconstruction errors of axial forces and bending moments to comprehensively evaluate the suitability of measurement points. Taking a tunnel in Anhui as an example， the study comprehensively analyzed the impact of factors such as the rock pressure sharing ratio， the horizontal rock pressure coefficient， and measurement noise on the estimation of safety factors. The estimated safety factors generally followed a normal distribution trend， with larger fluctuations at the mid-arch and arch shoulder positions. Aiming to reduce estimation uncertainties and biases in the reconstruction of internal force information， measurement points were arranged in accordance with on-site working conditions. Long-term monitoring data indicated that the internal forces in the secondary lining structure fluctuated significantly due to various factors at the initial stage of casting and stabilized after 100 days， ensuring an adequate safety reserve for the structure. This research can provide references for the statistical analysis of safety factors and sensor placement in tunnel lining structures.

In order to realize the high-value and large-scale utilization of metallurgical solid waste materials， the soil curing agent was synergistically prepared by using metallurgical solid waste materials （blast furnace slag， fly ash， calcium carbide slag and gypsum）， and the effects of different solid waste materials on the mechanical properties of the cured soil were investigated by orthogonal ratio optimization experiments and the optimal parameter formulations were obtained. The mechanical properties （unconfined compressive strength， splitting strength， delayed molding） and weathering properties （water stability， freeze-thaw cycle） of the solidified soil with solid waste-based curing agent （SWC） were systematically studied and compared with PO 42.5 silicate cement （OPC）. The results showed that： calcium carbide slag played an important role in the curing agent system， and the appropriate gypsum dosage had a positive effect on the strength growth； the strength of cured soil increased with the increase of SWC dosage and the age of curing， and the 7 day unconfined compressive strength of the cured soil at 4% dosage was more than 1 MPa， and the splitting strength was basically comparable to that of the OPC cured soil； the SWC cured soil showed a longer allowable delay in construction， and there was no trend of decreasing strength within 12 h after mixing； the water stability coefficient was more than 90% when the dosage was more than 5%， and the anti-freezing coefficient under the same dosage was increased by 3%-5% compared with that of cement， so it had good water stability and anti-freezing property； X-ray Diffraction （XRD） and Scanning Electron Microscope （SEM） analyses showed that SWC cured soils generate expansionary hydration products， calcite alumina （AFt） and hydrated calcium silicate （C-S-H）， during the hydration process， and that calcite crystals played a significant role in improving the cleavage strength of cured soils and increasing the resistance to cracking and deformation.

In order to lessen the environmental issues caused by the accumulation of titanium gypsum and to also alleviate the current situation of road material shortage， titanium gypsum and silane coupling agent-treated titanium gypsum were used to partially or completely replace limestone mineral powder. Additionally， the effects of titanium gypsum and modified titanium gypsum on the conventional and rheological properties of asphalt mastic were investigated. First， a laser particle size meter and a scanning electron microscope were used to examine the micro-morphology and particle size distribution of titanium gypsum， modified titanium gypsum， and limestone mineral powder. Subsequently， asphalt mortar containing varying amounts of titanium gypsum and modified titanium gypsum dosage was prepared. The asphalt mortar's three main indicators （penetration， ductility， and softening point）， as well as its high temperature and low temperature rheological performances， fatigue resistance， were tested using conventional testing methods， temperature scanning test， frequency scanning test， multiple stress creep recovery test （MSCR）， linear amplitude scanning test （LAS）， and bending beam rheological test （BBR）. The findings indicated that： titanium gypsum and modified titanium gypsum had a wider particle size range and a rougher surface. The addition of titanium gypsum and modified titanium gypsum decreased the asphalt mortar's ductility， penetration， unrecoverable creep flexibility （Jnr）， fatigue life （N _f）， creep rate （m） and phase angle （δ）， while the softening point， complex modulus （G*）， resilience （R） and stiffness modulus （S） were increased. In other words， performance at high temperatures was enhanced while low temperatures and weariness were decreased. The low-temperature performance was decreased and the high-temperature performance and elastic recovery indexes of modified titanium gypsum outperformed those of titanium gypsum at the same temperature. The determination of the viability of utilizing titanium gypsum in the creation of asphalt mixtures provides a path forward for the resource's exploitation.

Based on the large-span asymmetric small-spacing tunnel project of Pingnan Road in Longgang District of Shenzhen， the internal force changes of tunnel lining are analyzed through on-site monitoring results. A finite element model is established to compare the arch settlement， the stability of the middle rock pillar and the lining stress of the tunnel under different construction sequences. The results show that， the side arch waist and arch foot near the middle rock pillar are the structural instability points during the construction of large span asymmetric small spacing tunnel， and reinforcement treatment is needed in the construction. Compared with the large section first， when the small section of the tunnel is first， the arch settlement， lining bending moment and axial force of the large and small sections are reduced by up to 16.1%， 10.6% and 12.5%， respectively， and the stability of the middle rock pillar is improved. The horizontal stress and vertical displacement are reduced by up to 16.7% and 28.3%， respectively， which is beneficial to the overall safety of the structure. Combined with the comparative analysis of field monitoring data and numerical results， the maximum errors of axial force and bending moment are controlled within 12.5% and 16.5%， respectively. The overall trend of numerical results is basically consistent with the field monitoring results， which verifies the accuracy and feasibility of the calculation model. Therefore， it is reasonable to follow the principle of ‘small section first， large section backward， and backward hole far away from the middle rock pillar area first’ in the construction of large span asymmetric small spacing tunnel.

To improve the accuracy of non-destructive estimation of volume of trees with complex trunk shapes， this study uses terrestrial laser scanning （TLS） point cloud data and developes a taper-binary volume equation based on the optimal taper model for plantation Mongolian oak. The study was conducted in a Mongolian oak plantation located at the Urban Forestry Demonstration Base in Harbin. Complete point cloud data were collected via TLS， and tree trunk structural parameters were extracted after processes such as clipping， elevation normalization， filtering， individual tree segmentation， and leaf-branch separation. Based on the trunk shape characteristics of Mongolian oak， six taper equation models were tested （Biging（1984）， Amidon（1984）， Meng Xianyu（1982）， Kozak（2004）-Ⅱ， Zeng Weisheng et al.（1997）， Max and Burkhart（1976））， and the best-fitting model was selected through nonlinear regression to construct the taper-binary volume equation. Results showed that the accuracy of individual tree identification was 95.22%， with the coefficient of determination （R²） for extracted tree height and diameter at breast height （DBH） being 0.97 and 0.98， respectively， when compared to field measurements. The optimal taper model achieved an R² of 0.99 and a root mean square error （RMSE） of 0.38 cm. Residual analysis of the Mongolian oak taper-binary volume equation from this study with the existing volume equation showed the reliability of its estimation results. It can provide important technical support for estimating the volume of trees with complex trunk shapes using TLS point cloud data.

To address the issue that traditional semi-supervised self-training classification methods can lead to dataset confusion， affecting the accuracy of subsequent small-sample tree species classification， an EW-EL（entropy weight and ensemble learning） semi-supervised small-sample tree species classification method is proposed based on the entropy weight method（EW） and ensemble learning（EL）. EW-EL introduces the concept of EL into the theoretical framework of traditional semi-supervised self-training classification methods， using the entropy weight method as a foundational theory. It calculates the information entropy based on the F1 score of base classifiers in the current training cycle as a weight factor. Then， design the weights according to the idea that the larger the information entropy， the more unstabel the base classifier will be. This will make the classification probabilities of the ensemble classifier more concentrated and reduce the bias of the ensemble classifier. The findings demonstrate that， in contrast to conventional semi-supervised self-training techniques， EW-EL can efficiently balance data distribution， producing more precise pseudo-label sample categories for recently added data. With a recall of 0.96 and a Kappa coefficient of 0.97， the overall accuracy（OA） of the EW-EL method for small-sample tree species classification is 0.97. All three indicators are superior to supervised classification， conventional semi-supervised self-training techniques， and semi-supervised self-training techniques built using conventional EL mechanisms. In particular， the EW-EL approach outperforms semi-supervised self-training techniques that incorporate a soft voting mechanism in terms of OA and recall by 1%. Furthermore， in the chosen test area， the tree species map produced with EW-EL in combination with basic linear iterative clustering reached 94% accuracy. Moreover， extra analyses show that EW-EL can integrate several classifiers to provide better small-sample tree species classification results， which makes it more appropriate for relevant departments in forestry resource statistics under low-cost circumstances.

In order to solve the problem that the target detection algorithm is prone to leakage detection and insufficient detection accuracy in pedestrian detection in forest areas， a forest pedestrian target detection algorithm based on improved YOLOv8 is proposed. The C2f_DWRSeg module is used to replace the C2f module， and the number of initial convolutional channels is expanded so that the network can extract multi-scale features more efficiently. A reconstructed detector head is proposed to increase the complexity of the convolution layer during training， and a single branch structure is used in inference， so as to enrich the feature representation of the network and maintain efficient inference speed； add CGA， the convolution attention mechanism module， before feature fusion， to reduce the amount of calculation； use the Focaler-ShapeIoU loss function to replace the CIoU loss function to make up for the shortcomings of the boundary box regression method and further improve the detection ability. Experimental results show that compared with benchmark model， the improved algorithm mAP50 has increased by 2%， mAP50-95 has increased by 2.4%， and FPS has increased by 4.33%. It proves that the improved algorithm can be better applied to the task of pedestrian detection in forest areas.

Dioryctria is a kind of pest that seriously harms conifer species and seriously affects the health and growth of conifer trees. The larvae of Dioryctria feed on the leaves of coniferous trees and build nests in coniferous trees. Gradually， they destroy the leaf tissues， causing the needles to turn yellow and eventually leading to the withering of the trees. In addition， larvae may also erode the bark of trees， resulting in bark flaking and trunk exposure， leaving trees vulnerable to other pests， germs and natural elements， increasing the vulnerability of trees and reducing their viability. In order to assist the ground treatment of trees eaten by Dioryctria， the YOLOv8s target detection algorithm was adopted to realize the detection and recognition of the trees eaten by Dioryctria. By using C2f-GAM and dynamic detection head to build a model （YOLOv8-DM）， the detection ability of YOLOv8s against Dioryctria moth trees was improved. The experimental results showed that YOLOv8-DM could effectively identify Dioryctria moth trees with an average accuracy of 84.8%. Compared with other target detection algorithms， YOLOv8-DM has higher average precision.

Studying the spatial-temporal evolution and driving factors of forest vegetation in the three parallel rivers region can provide data support for building the ecological safety barrier and promoting the social and economic development in Southwest China. The spatial-temporal evolution pattern of forest vegetation in the three parallel rivers region was explored by geospatial analysis， and the evaluation index system was constructed from the two dimensions of ‘natural-social economy’. And 14 specific evaluation indexes （elevation， slope and annual precipitation， etc.） were selected to measure the driving force of the area change of each vegetation type in the study area. The results were as follows， 1） there were three types of forest vegetation in the study area： evergreen broad-leaved forest， deciduous broad-leaved forest and evergreen needle-leaved forest， among which evergreen needle-leaved forest accounted for more than 84% of the forest vegetation. 2） There were three change track of stablility， mutation and fluctuation， among which the stablility type accounted for 80.35% of the changing area. The spatial distribution of “cold-hot spots” in evergreen broad-leaved forest and deciduous broad-leaved forest was similar， and the “cold-hot spots” in evergreen needle-leaved forest showed a trend of cold spots surrounding hot spots. 3） The factor detection showed that the influence of each factor on evergreen broad-leaved forest was greater than others， and the explanation strength for evergreen needle-leaved forest was small. The spatial distribution of forest vegetation in the three parallel rivers region has obvious spatial heterogeneity， and the interaction effect of elevation and other factors on forest vegetation in the study area is significantly stronger than that of single factor detection.