Computer vision-based precision weed control has proven effective in reducing herbicide usage, lowering weed management costs, and enhancing sustainability in modern agriculture. However, developing deep learning models remains challenging due to the effort required for weed dataset annotation and the difficulty of identifying weeds at different stages and densities in complex field conditions. To address these challenges, this study introduces an indirect weed detection method that combines deep learning and image processing techniques. The proposed approach first employs an object detection network to identify and label crops within the images. Subsequently, image processing techniques are applied to segment the remaining green pixels, thereby enabling indirect detection of weeds. Furthermore, a novel detection network–CD-YOLOv10n (You Only Look Once version 10 nano)–was developed based on the YOLOv10 framework to optimize computational efficiency. By redesigning the backbone (C2f-DBB) and integrating an optimized upsampling module (DySample), the network achieved higher detection accuracy while maintaining a lightweight structure. Specifically, the model achieved a mean average precision (mAP50) of 98.1%, which is a 1.4% percentage-point increase compared with the YOLOv10n baseline, a relevant improvement given the already strong baseline performance. At the same time, compared to YOLOv10n, its GFLOPs were reduced by 22.62%, and the number of parameters decreased by 15.87%. These innovations make CD-YOLOv10n highly suitable for deployment on resource-constrained platforms.

To investigate multiple effects of the interaction between V. cholerae and phage on cholera transmission, we propose a degenerate reaction-diffusion model with different dispersal rates, which incorporates a short-lived hyperinfectious (HI vibrios) state of V. cholerae and lower-infectious (LI vibrios) state of V. cholerae. Our main purpose is to investigate the existence and stability analysis of multi-class boundary steady states, which is much more complicated and challenging than the case when the boundary steady state is unique. In a spatially heterogeneous case, the basic reproduction number $\mathscr{R}_{0}$ is defined as the spectral radius of the sum of two linear operators associated with HI vibrios infection and LI vibrios infection. If $\mathscr{R}_{0}\leq 1$, the disease-free steady state is globally asymptotically stable. If $\mathscr{R}_{0}\gt 1$, the uniform persistence of phage-free model, as well as the existence of the phage-free steady state, are established. In a spatially homogeneous case, when $\ \;\widetilde{\!\!\!\mathscr{R}}_{0}\gt 1$, the global asymptotic stability of phage-free steady state and the uniform persistence of the phage-present model are discussed under some additional conditions. The mathematical approach here has wide applications in degenerate Partial Differential Equations.

In small-plot experiments, weed scientists have traditionally estimated herbicide efficacy through visual assessments or manual counts with wooden frames—methods that are time-consuming, labor-intensive, and error-prone. This study introduces a novel mobile application (app) powered by convolutional neural networks (CNNs) to automate the evaluation of weed coverage in turfgrass. The mobile app automatically segments input images into 10 by 10 grid cells. A comparative analysis of EfficientNet, MobileNetV3, MobileOne, ResNet, ResNeXt, ShuffleNetV1, and ShuffleNetV2 was conducted to identify weed-infested grid cells and calculate weed coverage in bahiagrass (Paspalum notatum Flueggé), dormant bermudagrass [Cynodon dactylon (L.) Pers.], and perennial ryegrass (Lolium perenne L.). Results showed that EfficientNet and MobileOne outperformed other models in detecting weeds growing in bahiagrass, achieving an F1 score of 0.988. For dormant bermudagrass, ResNet performed best, with an F1 score of 0.996. Additionally, app-based coverage estimates (11%) were highly consistent with manual assessments (11%), showing no significant difference (P = 0.3560). Similarly, ResNeXt achieved the highest F1 score of 0.996 for detecting weeds growing in perennial ryegrass, with app-based and manual coverage estimates also closely aligned at 10% (P = 0.1340). High F1 scores across all turfgrass types demonstrate the models’ ability to accurately replicate manual assessments, which is essential for herbicide efficacy trials requiring precise weed coverage data. Moreover, the time for weed assessment was compared, revealing that manual counting with 10 by 10 wooden frames took an average of 39.25, 37.25, and 42.25 s per instance for bahiagrass, dormant bermudagrass, and perennial ryegrass, respectively, whereas the app-based approach reduced the assessment times to 8.23, 7.75, and 14.96 s, respectively. These results highlight the potential of deep learning–based mobile tools for fast, accurate, scalable weed coverage assessments, enabling efficient herbicide trials and offering labor and cost savings for researchers and turfgrass managers.

The Early-Middle Jurassic impression/compression macroflora and the palynoflora from the Qaidam Basin in the northeastern Qinghai-Xizang (Tibetan) Plateau have been well studied; however, fossil wood from this region has not been previously documented systematically. Here, we describe an anatomically well-preserved fossil wood specimen from the Lower Jurassic Huoshaoshan Formation at the Dameigou section in northern Qinghai Province, northwestern China. This fossil exhibits typical Metapodocarpoxylon Dupéron-Laudoueneix et Pons anatomy with usually araucarian radial tracheid pits and variable cross-field pits, representing a new record for Metapodocarpoxylon in the Qaidam Basin. This discovery indicates that trees with this type of wood anatomy were not confined to northern Gondwana but also grew in more northerly regions in Laurasia. The wood displays distinct growth rings, with abundant, well-formed earlywood and narrow latewood. This observation, along with previous interpretations based on macroflora, palynoflora and sedimentological data, suggests that a warm and humid climate with mild seasonality prevailed in the region during the Early Jurassic.

Machine vision–based herbicide applications relying on object detection or image classification deep convolutional neural networks (DCNNs) demand high memory and computational resources, resulting in lengthy inference times. To tackle these challenges, this study assessed the effectiveness of three teacher models, each trained on datasets of varying sizes, including D-20k (comprising 10,000 true-positive and true-negative images) and D-10k (comprising 5,000 true-positive and true-negative images). Additionally, knowledge distillation was performed on their corresponding student models across a range of temperature settings. After the process of student–teacher learning, the parameters of all student models were reduced. ResNet18 not only achieved higher accuracy (ACC ≥ 0.989) but also maintained higher frames per second (FPS ≥ 742.9) under its optimal temperature condition (T = 1). Overall, the results suggest that employing knowledge distillation in the machine vision models enabled accurate and reliable weed detection in turf while reducing the need for extensive computational resources, thereby facilitating real-time weed detection and contributing to the development of smart, machine vision–based sprayers.

Two-dimensional oblique detonation wave (ODW) propagations in partially prevaporized n-heptane sprays are numerically simulated with a skeletal chemical mechanism. The influences of the droplet diameter and total equivalence on oblique detonation are considered. The initiation length is found to increase first and then decrease with increasing initial droplet diameter, and the effect of droplet size is maximized when the initial droplet diameter is approximately $10\ \mathrm {\mu } {\rm m}$. As the initial droplet diameter varies, unsteady and steady ODWs are observed. In the cases of unsteady ODWs, temperature gradients and non-uniform distributions of the reactant mixture due to droplet evaporation lead to formation of unsteady detonation propagation, therefore leading to fluctuations in the initiation length. The fluctuations in initiation length decrease as the pre-evaporation gas equivalence ratio increases for the unsteady cases. The results further suggest that the relationship between the evaporation layer thickness along the streamline and the corresponding theoretical initiation length can be used to identify an unsteady or steady ODW in cases with large droplets that evaporate behind an oblique shock wave or ODW under the effects of different initial droplet diameters.

In the present study, acid-modified attapulgite was used, as an adsorbent, to remove as much Cd2+ as possible from aqueous solution. Static adsorption experiments using powdered acid-modified attapulgite, and dynamic adsorption using granular acid-modifed attapulgite, were conducted to explore the practical application of modified attapulgite in the adsorption of Cd2+. The modified attapulgite had a larger specific surface area and thinner fibrous crystals than the unmodified version. No obvious differences were noted, in terms of the crystal structure, between the natural attapulgite and the modified version. The effects of initial concentration, pH, contact time, and ionic strength on the adsorption of Cd2+ were investigated, and the results showed that the adsorption capacity of the modified attapulgite was increased with increasing pH and the initial Cd2+ concentration. The adsorption properties were analyzed by means of dynamic adsorption tests with respect to various Cd2+ concentrations and flow rates. The maximum adsorption capacity of 8.83 mg/g occurred at a flow rate of 1 mL/min and at an initial concentration of 75 mg/L. Because there was better accord between the data and a pseudo-second order model than a pseudo-first-order model, external mass transfer is suggested to be the rate-controlling process. The experimental data were also fitted for the intraparticle diffusion model, implying that the intraparticle diffusion of Cd2+ onto the modified attapulgite was also important for controlling the adsorption process. The Bohart-Adams model was more suitable than the Thomas model for describing the dynamic behavior with respect to the flow rate and the initial Cd2+ concentration. This research provided the theoretical basis for the dynamic adsorption of Cd2+ on the modified attapulgite. Compared to the powdered modified attapulgite, the dynamic adsorption by granular modified attapulgite appeared more favorable in terms of practical application.

Dietary antioxidant indices (DAI) may be potentially associated with relative telomere length (RTL) of leucocytes. This study aimed to investigate the relationship between DAI and RTL. A cross-sectional study involving 1656 participants was conducted. A generalised linear regression model and a restricted cubic spline model were used to assess the correlation of DAI and its components with RTL. Generalised linear regression analysis revealed that DAI (β = 0·005, P = 0·002) and the intake of its constituents vitamin C (β = 0·043, P = 0·027), vitamin E (β = 0·088, P < 0·001), Se (β = 0·075, P = 0·003), and Zn (β = 0·075, P = 0·023) were significantly and positively correlated with RTL. Sex-stratified analysis showed that DAI (β = 0·006, P = 0·005) and its constituents vitamin E (β = 0·083, P = 0·012), Se (β = 0·093, P = 0·006), and Zn (β = 0·092, P = 0·034) were significantly and positively correlated with RTL among females. Meanwhile, among males, only vitamin E intake (β = 0·089, P = 0·013) was significantly and positively associated with RTL. Restricted cubic spline analysis revealed linear positive associations between DAI and its constituents’ (vitamin E, Se and Zn) intake and RTL in the total population. Sex-stratified analysis revealed a linear positive correlation between DAI and its constituents’ (vitamin E, Se and Zn) intake and RTL in females. Our study found a significant positive correlation between DAI and RTL, with sex differences.

This paper is concerned with a nonlocal reaction–diffusion system with double free boundaries and two time delays. The free boundary problem describes the evolution of faecally–orally transmitted diseases. We first show the well-posedness of global solution, and then establish the monotonicity and asymptotic property of basic reproduction number for the epidemic model without delays, which is defined by spectral radius of the next infection operator. By introducing the generalized principal eigenvalue defined in general domain, we obtain an upper bound of the limit value of basic reproduction number. We discuss the spreading and vanishing phenomena in terms of the basic production number. By employing the perturbed approximation method and monotone iteration method, we establish the existence, uniqueness and monotonicity of solution to semi-wave problem. When spreading occurs, we determine the asymptotic spreading speeds of free boundaries by constructing suitable upper and lower solutions from the semi-wave solutions. Moreover, spreading speeds for partially degenerate diffusion case are provided in a similar way.

A delayed reaction-diffusion system with free boundaries is investigated in this paper to understand how the bacteria spread spatially to larger area from the initial infected habitat. Under the assumptions that the nonlinearities are of monostable type and the initial values satisfy some compatible condition, we show that the free boundary problem is well-posed and discuss the long-time behaviour of solution (including spreading and vanishing) in terms of the spatial-temporal risk index. Furthermore, to determine the spreading speed of free boundaries when spreading occurs, we first study the distribution of roots of a transcendental equation containing a polynomial of degree four and then establish the existence and uniqueness of monotone solution to a delay-induced nonlocal semi-wave problem by employing the approximation method, lower-upper solutions technique and Schauder fixed point theorem. It is shown that time delays slow down the spreading of bacteria.

Previous hydrocarbon explorations in the middle of the Tarim Basin indicate that strike-slip faults play an important role in the development of Ordovician carbonate reservoirs and hydrocarbon accumulation. The SB5 fault in the Tarim Basin was the target of this investigation. An evaluation of the stress in situ was carried out and provided boundary conditions to build a 3D geomechanical model. The distribution and application of present in situ stress in the strike-slip fault were studied. The results show good agreement between the absolute measured stress in situ and the modelled stresses, revealing a different stress regime along the strike-slip fault. The uplift segment belongs to a strike-slip stress state, and other areas belong to a normal fault stress state. The strike-slip fault has a significant influence on the present in situ stress distribution. The direction of the maximum horizontal stress deflects near the fault and tends to be parallel to the fault strike. This work introduces a comprehensive evaluation of the present in situ stress of the fractured carbonate reservoirs controlled by the strike-slip fault system. The present in situ stress direction can clarify the propagation direction of hydraulic fracturing and serve to evaluate the effectiveness of natural fractures.

The Huangshaping deposit is unique in southern Hunan Province, China, as it hosts economic reserves of both W–Mo and Pb–Zn mineralization, which are usually associated with granite and granodiorite porphyry in this area, respectively. This study reports results of in situ LA-MC-ICP-MS sulphur isotopic composition analyses conducted on sulphides from both W–Mo and Pb–Zn mineralization from the Huangshaping deposit with the aim of constraining ore genesis for this deposit. All samples from the proximal W–Mo mineralization have relatively uniform and high δ34S values (8.7 ‰ to 16.0 ‰), close to the range for carbonate sediments in this deposit (13.8 ‰ to 18.1 ‰). These patterns suggest that the granite porphyry in this deposit was the sulphur source for W–Mo mineralization, and that the assimilation of evaporite from the carbonate sediments led to the high δ34S values of the granite porphyry. Sulphides from the Pb–Zn mineralization have δ34S values (2.2 ‰ to 10.3 ‰) lower than those of the W–Mo mineralization, and generally increase in this paragenetic order, with the lowest δ34S values being similar to those of the basement (3.8 ‰ to 7.7 ‰). These patterns indicate that the original sulphur for the Pb–Zn mineralization was most likely derived from the basement, with input of sulphur from the carbonate sediments increasing during the evolution of ore-forming fluids. On the basis of the measured sulphur isotopic compositions, it is suggested that the ore-forming materials for the W–Mo mineralization were derived from the granite porphyry, whereas ore-forming materials extracted from the basement dominated the Pb–Zn mineralization.

Manure is a primary source of methane (CH4) emissions into the atmosphere. A large proportion of CH4 from manure is emitted during storage, but this varies with storage methods. In this research, we tested whether covering a manure heap with plastic reduces CH4 emission during a short-term composting process. A static chamber method was used to detect the CH4 emission rate and the change of the physicochemical properties of cattle manure which was stored either uncovered (treatment UNCOVERED) or covered with plastic (treatment COVERED) for 30-day periods during the four seasons? The dry matter content of the COVERED treatment was significantly less than the UNCOVERED treatment (P < 0.01), and the C/N ratio of the COVERED treatment significantly greater than the UNCOVERED treatment (P > 0.05) under high temperature. In the UNCOVERED treatment, average daily methane (CH4) emissions were in the order summer > spring > autumn > winter. CH4 emissions were positively correlated with the temperature (R2 = 0.52, P < 0.01). Compared to the UNCOVERED treatment, the daily average CH4 emission rates from COVERED treatment manure were less in the first 19 days of spring, 13 days of summer, 10 days of autumn and 30 days of winter. In summary, covering the manure pile with plastic reduces the evaporation of water during storage; and in winter, long-term covering with plastic film reduces the CH4 emissions during the storage of manure.

With the characteristics of full autonomy and no accumulated errors, polarisation navigation shows tremendous prospects in underwater scenarios. In this paper, inspired by the polarisation vision of aquatic organisms, a novel point-source polarisation sensor with high spectral adaptability (400 nm–760 nm) is designed for underwater orientation. To enhance the environmental applicability of the underwater polarisation sensor, a novel sensor model based on the underwater light intensity attenuation coefficient and optical coupling coefficient is established. In addition, concerned with the influence of light intensity uncertainty on sensor performance underwater, an antagonistic polarisation algorithm is adopted for the first time, to improve the accuracy of angle of polarisation and degree of polarisation in the low signal-to-noise ratio environment underwater. Finally, indoor and outdoor experiments are carried out to evaluate the performance of the designed polarisation sensor. The results show that the designed point-source polarisation sensor can acquire polarised light and be used for heading determination underwater.

Interscale kinetic energy transfer in chemically reacting compressible isotropic turbulence is studied using numerical simulations at turbulent Mach numbers 0.2 and 0.8 for isothermal and exothermic reactions. At low turbulent Mach number $M_{t}=0.2$ for exothermic reaction, heat release greatly enhances expansion and compression motions, and induces the formation of shocklets which are not observed for isothermal reaction at the same turbulent Mach number. It is found that heat release through exothermic reactions enhances both the positive and the negative components of pressure–dilatation, as well as positive and negative components of subgrid-scale (SGS) kinetic energy flux, indicating an increase of both forward-scatter and backscatter of kinetic energy. The SGS flux of kinetic energy has a tendency to be from large scales to small scales. The solenoidal components of pressure–dilatation and SGS kinetic energy flux are minorly influenced by heat release of reaction, while the dilatational components of pressure–dilatation and SGS kinetic energy flux are significantly intensified by heat release at all length scales. Heat release enhances the kinetic energy of the dilatational mode through the pressure work, and leads to the increase of dilatational kinetic energy transfer from large scales to small scales through dilatational SGS flux, as well as the viscous dissipation of dilatational kinetic energy at small scales. Taylor Reynolds number has a minor influence on the qualitative statistical properties of interscale kinetic energy transfer terms.

The Western Tianshan orogenic belt is essential for understanding the evolution of the Central Asian orogenic belt. However, no agreement exists among geologists about its tectonic environment during the Late Palaeozoic. The volcanic rocks of the Yishijilike and Wulang Formation in the Yili Block, Western Tianshan, formed in the Late Carboniferous to Early Permian, mainly consist of a bimodal suite of basalts – basaltic andesites and rhyolites, with only some intermediate rocks. Mafic rocks are slightly enriched in light rare earth elements (LREE) and depleted in Nb, Ta, Zr and Hf, suggesting a subduction-modified depleted mantle source. Some mafic samples in the Early Permian bimodal volcanic rocks have high Ti contents with relatively high concentrations of Nb and high field strength elements (HFSE) and low contents of heavy rare earth elements (HREE). These rocks are similar to the continental flood basalts, which suggests that they formed from an asthenospheric mantle. This paper indicates that mafic members were created by the partial melting of the asthenospheric mantle material and subduction-modified lithospheric mantle mixture. Some rhyolites and dacites in the Wulang formation were enriched in Ga, Nb, Zr, Ce and Y and depleted in Sr and Eu. Additionally, they showed fractionation of rare earth elements (REE) with negative Eu anomalies, which is indicative of an A-type affinity of felsic rocks. The genesis of mafic members and an A-type affinity of felsic members indicate that the Late Carboniferous – Early Permian magmatism in the Western Tianshan area formed as a result of an extensional setting. This study also reveals bimodal magmatism produced by delamination in an extensional tectonic setting.