Changes in tropical cyclone (TC) intensity are a critical determinant of their destructive potential. However，our understanding of TC weakening processes remains limited.This paper systematically reviews the environmental factors and physical mechanisms governing TC intensity decay. First，the definition of TC weakening and the differences in rapid weakening definitions across ocean basins are presented. Then，the major environmental factors responsible for TC weakening and the associated physical mechanisms are overviewed. The ocean reduces the energy supply by decreasing sea surface temperatures (SST)，enhancing horizontal SST gradients，and triggering cold wake feedback mechanisms. Vertical wind shear disrupts the warm⁃core structure by inducing vortex tilt and ventilation effects. Land surfaces accelerate weakening by increasing surface roughness to enhance frictional dissipation and by cutting off energy sources through drying processes. Aerosols modulate TC intensity changes through aerosol⁃radiation interactions and aerosol⁃cloud interactions via multiple pathways. Multiscale weather systems，including upper⁃level troughs，monsoon systems，and binary TC interactions，trigger TC weakening by coupling multiple adverse factors. Finally，the long⁃term trends of TC weakening on climatic scales are summarized. The decay process of global major TCs has accelerated，and the TC weakening rate over the western North Pacific has increased. Future research needs to deepen the understanding of multiscale interactions and develop high⁃resolution coupled models integrated with artificial intelligence approaches to improve TC weakening forecast capability，thereby providing scientific support for disaster prevention and mitigation.

Quasi⁃two⁃dimensional convection⁃permitting simulations under an idealized radiative⁃convective equilibrium framework are used to examine how latitude,topography,and island size shape the spatiotemporal characteristics of tropical island precipitation. Daily mean precipitation decreases with increasing latitude,and a pronounced diurnal cycle appears,with rainfall peaking from afternoon to evening. The peak occurs earlier at higher latitudes (~14:00 LT) and shifts to the evening at lower latitudes (~21:00 LT),partly due to latitude⁃dependent variations in sea–land breeze circulations. Increasing topographic height enhances precipitation by strengthening vertical upwelling over the island's interior. Enlarging island size does not alter the overall circulation structure,but leads to more spatially dispersed rainfall,reducing the island⁃mean daily precipitation despite an increase in total convective precipitation.

Utilizing summertime multi⁃S⁃band operational radar mosaic data (2016-2018) over North China，GPM⁃IMERG (Integrated Multi⁃satellite Retrievals for Global Precipitation Measurement) satellite land⁃sea mask data，this study develops a set of objective identification criteria for offshore⁃propagating MCSs based on dynamic underlying surface properties. Building upon this，the statistical characteristics of offshore⁃propagating and non⁃offshore⁃propagating MCS are analyzed，and the north⁃south differences in the large⁃scale circulation background and environmental driving mechanisms for offshore⁃propagating MCS over North China are revealed. The principal conclusions are as follows.(1) The established objective identification criteria effectively capture the dynamic process of MCS crossing the coastline. Statistics show that offshore⁃propagating MCS samples are characterized by being "less frequent but more intense". (2) During the occurrence of offshore⁃propagating MCSs，North China is located within a strong upper⁃level divergence area to the right of the 200 hPa jet entrance region. The mid⁃to⁃lower troposphere exhibits stronger baroclinicity，and the environmental fields feature significantly higher Most Unstable Convective Available Potential Energy (MUCAPE) and vertical wind shear，providing favorable conditions for the organization and maintenance of convection. (3) Significant north⁃south differences exist in the driving mechanisms for MCS offshore propagation over North China: The northern Bohai Sea region belongs to the "dynamic guidance type"， where systems are steered southeastward into the sea by a 500 hPa shortwave trough. Post⁃trough dry，cold air advection at 700 hPa enhances convective instability，and strong vertical wind shear helps balance the cold pool outflow，sustaining the system's offshore propagation. In contrast，the southern Bohai Sea region is characterized as the "warm⁃moist advection type，" driven by a robust 700 hPa southwesterly low⁃level jet (LLJ). The LLJ not only provides steering but also continuously transports energy and moisture，supporting the system's northeastward extension within a warm and moist background.This research not only advances the physical understanding of convective system evolution across the land⁃sea transition zone but also provides a technical foundation and new perspectives for future studies on the mechanisms governing MCS offshore evolution. Furthermore，it offers objective guidance for the operational forecasting of severe convective weather in coastal regions.

Under the influence of the weakened low⁃pressure circulation of Typhoon "IN⁃FA" after landfall，a narrow convective rainband developed in Tianchang area along the border of Jiangsu and Anhui provinces in the afternoon of July 28，2021. The rainband moved in the direction of east by north，causing severe wind and rainfall in Tianchang and surrounding areas. Based on multi⁃source observations and high⁃resolution simulations using the Weather Research and Forecasting (WRF) model，this study investigates the formation mechanisms of the convective rainband from both dynamic and thermodynamic perspectives. Results indicate that the typhoon's low⁃pressure circulation provided a warm and humid environment favorable for convection. Afternoon solar radiation increased surface temperatures，enabling the near⁃saturated air to hold more moisture. Combined with water vapor transport by southerly flow on the eastern flank of the low⁃pressure circulation，low⁃level humidity and convective available potential energy (CAPE) increased，leading to the development of convective instability. The interaction between the low⁃pressure circulation and the westerly trough introduced dry and cold air in the mid⁃troposphere，further enhancing instability through an upper⁃dry and lower⁃wet stratification. Perturbation convergence of low⁃level winds triggered local convection，while convergence in the middle and lower levels and upward vorticity transport contributed to local vorticity growth. After initiation，convective cells organized into a north⁃south⁃oriented severe convective rainband that propagated east⁃northeastward under the influence of the low⁃pressure circulation. Sensitivity experiments highlight the critical role of afternoon radiative heating in increasing low⁃level humidity and CAPE，thereby facilitating the development of the severe convective rainband.

This study develops a tracking algorithm for Mesoscale Convective Systems (MCSs) preceding tropical cyclone (TC) formation in the Western North Pacific from 2000 to 2023. MCSs resulting in TC genesis are designated as developing MCSs (D⁃MCSs), whereas systems that do not contribute to genesis are classified as non⁃developing MCSs (ND⁃MCSs). It is found that D⁃MCSs are present in environments characterized by elevated sea surface temperatures (SST) and a more humid mid⁃troposphere. They are also associated with enhanced low⁃level positive vorticity and increased high⁃cloud coverage. D⁃MCSs feature a deeper vertical vorticity structure with well⁃organized low⁃level convergence and upper⁃level divergence,indicative of enhanced convective organization. In contrast, ND⁃MCSs exhibit a shallower vertical structure and lack closed circulations. Further statistical analysis of D⁃MCSs reveals that those with more merging events exhibit prolonged durations from MCS formation to tropical cyclone genesis,larger initial vortex scales at cyclone genesis,and attenuated peak intensities over the TC's lifetime. The disparity in merging frequency is intimately linked to the spatial distribution patterns of MCSs. Cases featuring elevated merging events exhibit spatial concentration in the western sector of the Western North Pacific,where MCSs manifest greater spatiotemporal frequency. The propagation of MCSs is primarily modulated by large⁃scale circulation patterns. An intensified South China Sea monsoon trough induces a northwestward migration tendency in western systems,whereas enhanced low⁃level easterly steering flows associated with an intensified and expansive North Pacific subtropical high guide eastern systems westward.

Typhoon Danas，the 4th typhoon in 2025，made landfall in Zhejiang Province and was weakened into a remnant vortex,and moved southwestward across Fujian into Guangdong. Under the influence of this vortex，Guangdong Province experienced a widespread,extreme rainstorm on July 10-11. The extensive extreme rainstorm blanketed across the Pearl River Delta on July 10，when the center of the remnant vortex was over Eastern Guangdong. However,the torrential rains began to subside on July 11,as the remnant vortex center moved closer to the region. Traditional numerical models exhibited significant deviations in forecasting the location and temporal evolution of this heavy rainfall. To address the uncertainty in the typhoon rainstorm forecast，objective correction algorithms,such as frequency bias correction and the neighborhood method were applied to optimize the Bayesian model averaging (BMA). By leveraging the strengths of both artificial intelligence and traditional numerical models，a novel probabilistic precipitation forecast is developed based on the combination of the multiple forecasts. The probability of torrential rain predicted by the optimized BMA scheme successfully captured the major rainstorm area from the Pearl River Estuary to eastern Guangdong on July 10. On July 11, the likelihood of torrential rain decreased over the Pearl River Delta. The spatial distribution and temporal evolution of heavy rain probabilities exhibited better alignment with observations,effectively addressing the limitations of deterministic forecasts. A systematic verification of typhoon precipitation affecting Guangdong in 2025 shows that the optimized probabilistic precipitation forecast outperforms the ECMWF ensemble forecast in all verification metrics. These findings could serve as a valuable scientific basis for typhoon rainstorm prevention,mitigation,and emergency decision⁃making.

Based on daily precipitation data from 11 national surface meteorological stations (1991–2020), this study utilized the RClimDex model to calculate six precipitation indices and five precipitation⁃day indices,systematically investigating the evolution of extreme precipitation in the Ili River Basin, Xinjiang. Using linear trend analysis,the Mann⁃Kendall abrupt change test,and Morlet wavelet analysis,we systematically examined the spatiotemporal variation characteristics of extreme precipitation in the Ili River Basin,Xinjiang. The results showed that,with the exception of consecutive dry days and consecutive wet days (which decreased at rates of 0.86 and 0.12 d·（10 a）^-1,espectively),all other precipitation indices exhibited an upward trend. There were multiple mutation points in each index，and the mutation time was not obvious. Spatially,the maximum one⁃day precipitation shared a similar pattern with the maximum five⁃day precipitation,with values increasing from the central⁃western part of the Ili River Basin toward the surrounding areas. Annual precipitation decreased from the east and southwest toward the northwest of the basin. In contrast,the number of days with moderate,heavy,and torrential rain increased from the northwest toward the southeast. The dominant period for consecutive dry days was 8~12 a,with the oscillation being particularly strong during 2000-2010. The periodicity of consecutive wet days was weak,and short⁃term oscillations of 4~6 a were observed during certain periods. All other indices displayed significant short⁃period oscillations of 4~8 a over the 1991-2020 period. After 2005, the oscillation energy increased, with a corresponding intensification of interannual variation.

Post⁃training pruning (PTP) has emerged as an efficient compression technique to address the challenges of limited computational resources and excessive memory footprint during the edge deployment of Large Language Models (LLMs). However，existing mainstream methods (e.g.，Wanda and SparseGPT) typically employ uniform layer⁃wise sparsity strategies，overlooking the significant heterogeneity in information contribution across different layers and channels. Moreover，their evaluation criteria predominantly focus on input⁃side intensity，making it difficult to identify high⁃energy static redundant channels，which leads to severe model performance degradation under high compression ratios.To address these limitations，this paper proposes OGAS，an Output⁃activation Guided Adaptive Sparsity pruning method at the channel level. First，a dual evaluation metric is constructed by integrating the output activation energy norm with the Peak⁃to⁃Average Power Ratio (PAPR) to accurately identify and protect sparse key features from the dimensions of response intensity and feature specificity. Second，a continuous mapping mechanism based on non⁃linear curvature is designed to achieve dynamic adaptive allocation of channel⁃level sparsity within a continuous space. Furthermore，a closed⁃loop optimization workflow is established by introducing the Golden Section Search algorithm to realize the automated layer⁃wise tuning of critical hyperparameters.Experimental results on mainstream open⁃source models，including LLaMA⁃3 and Mistral，demonstrate that at a 50% sparsity ratio，OGAS reduces the perplexity (PPL) of LLaMA⁃3.1⁃8B on the WikiText⁃2 dataset to 7.99，a significant improvement over the state⁃of⁃the⁃art first⁃order method Wanda (8.85). In common sense reasoning tasks，the average zero⁃shot accuracy reached 63.46%，representing a 1.6% improvement over Wanda. The results verify that OGAS effectively maintains the semantic understanding and logical reasoning capabilities of models after large⁃scale compression，exhibiting superior robustness and versatility across different model architectures.

Open⁃wearable stereo (OWS) earphones have garnered widespread application due to their superior wearing comfort. However，the issue of sound leakage in quiet environments remains a critical challenge. Addressing the limitation that existing leakage models，most of which are derived from in⁃ear earphones⁃primarily focus on modeling the leakage gap while neglecting the influence of the pinna during propagation,this paper proposes a leakage model based on pinna measurements. This model decomposes the physical process of sound leakage into two distinct components：the equivalent sound source and the propagation path. First，an equivalent circuit model comprising two parallel branches⁃the open gap and structural vibration⁃is constructed to characterize the acoustic properties of sound leakage in OWS earphones. This reveals that far⁃field leakage is dominated by radiation from the open gap. Second，a Pinna⁃Related Transfer Function (PRTF) filter is introduced to characterize the path effects of leakage sound waves propagating around the pinna. Simulation and experimental results demonstrate that the proposed model more accurately fits the peak and valley characteristics of measured leakage data in the mid⁃to⁃high frequency range. This work provides a theoretical basis for the acoustic analysis of sound leakage and the structural design of leakage suppression in OWS earphones.

To address the issues of low efficiency and insufficient accuracy in multivariate long⁃sequence forecasting of PM_2.5 concentration time series data,a DAE⁃PatchTST⁃based PM_2.5 concentration prediction model is proposed.The data were collected from seven meteorological observation stations and atmospheric composition stations in different regions of Chongqing. Six sets of hourly historical meteorological and environmental data were selected using Pearson correlation coefficient analysis,and linear interpolation and reversible instance normalization methods were applied for data preprocessing. The constructed model utilizes a denoising autoencoder mechanism to reduce noise in the original input,thereby ensuring model stability. Meanwhile,channel independence and time series patch mechanisms were introduced into the classical Transformer architecture to extensively capture long⁃term dependencies in the input sequences. Comparative experiments with existing research methods such as RNN，GRU and LSTM demonstrate that the proposed model achieves the best performance in MAE，RMSE and R² metrics for short⁃and medium⁃term PM_2.5 concentration time series forecasting tasks. Additionally，the model exhibits strong reliability in predicting PM_2.5 concentrations in surrounding regions.

Novel physical state manipulation based on acoustic artificial structures has become a research hotspot in recent years. Among them，spoof surface acoustic waves (SSAWs) have garnered extensive attention owing to their surface⁃propagating characteristic and flexible modulatability by structures，exhibiting application potential in areas including high⁃resolution imaging and acoustic communication. However，the modulation approaches for SSAWs have limitations. In particular，effective phase encoding of SSAWs cannot be achieved currently，making it difficult to meet the requirements of practical application scenarios. Thus，there is an urgent need to explore new pathways for efficient modulation of SSAWs. This paper proposes a multifunctional coded acoustic metasurface based on Helmholtz resonators，achieving phase encoding of SSAWs by varying the cavity depth. Specifically，this paper designs an acoustic metalens that can achieve precise and tunable subwavelength acoustic focusing in a two⁃dimensional plane. In addition，this study further realizes the effective excitation of Airy beams by utilizing the asymmetric phase modulation mechanism，which is expected to have certain application potential in fields such as high⁃resolution medical ultrasonic imaging and acoustic communications.

Accurate sound field recording serves as a fundamental prerequisite for spatial audio applications such as sound field analysis，control，and reproduction. For directionally biased sound fields commonly encountered in practice，modal decomposition based on optimal basis functions has proven to be an efficient representation approach; however，existing implementations typically rely on independent measurements using a single microphone array，which provides limited spatial information and is susceptible to noise. To address this issue，this study incorporates multi⁃point measurement into the optimal modal decomposition framework and proposes a distributed collaborative recording method for directionally biased sound fields. By constructing a sound field decomposition transfer matrix under multi⁃point constraints using the addition theorem of spherical wave functions，joint estimation of modal decomposition coefficients is achieved. Numerical simulations and experimental measurements demonstrate that，compared with single⁃point strategies or methods using spherical harmonic bases，the proposed method significantly reduces the relative error of reconstructed sound pressure over both the effective frequency range and the entire target listening area，thereby improving recording performance for directionally biased sound fields.

To address the challenges of industrial pollution and fluctuations in environmental ammonia concentrations, ammonia sensors characterized by low cost,minimal power consumption,a wide detection range,and rapid response have garnered significant attention. In this study,a novel electrochemical sensor was fabricated via an MXene/WS₂ composite functionalized with polyethyleneimine (PEI)/polyethylene glycol (PEG). Characterization results demonstrate that the MXene serves as a two⁃dimensional conductive framework, while the WS₂ nanosheets provide abundant active sites. The PEI/PEG coating affords abundant -\mathrm{N}{\mathrm{H}}_{\mathrm{2}}/-OH functional groups,forming a P-N heterojunction structure that significantly enhances the material's gas sensing performance. Under a bias voltage of -0.3 V,the gas⁃sensing sensor demonstrated a response time of 17.2 s and a recovery time of 26.7 s toward 50 ppm ammonia. Ammonia chemisorbs on functional groups (e.g.,-\mathrm{N}{\mathrm{H}}_{\mathrm{2}} and Ti-(OH)ₓ) on the material's surface,forming adducts such as \mathrm{T}{\mathrm{i}}_{\mathrm{3}}{\mathrm{C}}_{\mathrm{2}}-\mathrm{N}{\mathrm{H}}_{\mathrm{3}}. The generation of these groups induces a change in the sensor's resistance. Experimental results demonstrate that the sensor exhibits excellent repeatability and stability (relative error1%),maintains stable accuracy at 25~35 °C and 25%~75% relative humidity,and shows good anti⁃interference ability against O₂,N₂,CO₂,and NO₂. This sensor is suitable for deep⁃well safety monitoring,indoor and outdoor air quality monitoring,and industrial exhaust gas monitoring.

Based on their different applications in integrated circuit manufacturing，semiconductor metal precursors are primarily divided into two categories: high⁃k precursors for high⁃k thin films and metal layer precursors for metal layer thin films. As Complementary Metal⁃Oxide⁃Semiconductor (CMOS) device feature sizes shrink below 45 nm，SiO₂ can no longer meet the performance and power consumption requirements as a gate dielectric material，making the replacement of SiO₂ with metal high⁃k materials an inevitable choice. Currently，HfO₂/SiO₂ stacks paired with a dipole layer LaO _x are predominantly used in logic circuits，while DRAM memory chips have transitioned from the previous ZrO₂/Al/ZrO₂ stack structure to a multicomponent system based on Zr _x Hf _y O，combined with other metal high⁃k stacks or dopants. However，an inherent trade⁃off of high⁃k dielectric materials is that a high dielectric constant is accompanied by a small band gap. Therefore，doping is necessary to form ternary or more complex multicomponent systems，leveraging the high⁃k value or high band gap characteristics of individual materials to achieve complementary performance. Additionally，as technology nodes shrink，the importance of advanced metal layers is increasingly prominent. For example，TaN is used as a barrier layer in back⁃end⁃of⁃line interconnects due to its excellent diffusion barrier properties and thermal stability. Mo，with its low resistivity and good scalability，has become a promising metal layer material in NAND memory chips. To meet the demands of advanced processes，metal layer precursors must possess high purity，high volatility，high stability，high reactivity，and good interfacial compatibility with adjacent materials. Based on past research achievements，this paper will summarize the high⁃k and metal precursors commonly used in existing process routes，along with related quality and packaging cylinder requirements. Furthermore，following the latest research and industrial roadmaps，it will provide some predictions regarding the future development directions of high⁃k and metal precursors for semiconductors，aiming to offer guidance for the existing industry and ultimately facilitate the development of novel metal high⁃k precursor materials suitable for industrialization.