In Earth system modeling,the land surface is coupled with the atmosphere through surface turbulent fluxes.These fluxes are computed using mean meteorological variables between the surface and a reference height in the...In Earth system modeling,the land surface is coupled with the atmosphere through surface turbulent fluxes.These fluxes are computed using mean meteorological variables between the surface and a reference height in the atmosphere.However,the dependence of flux computation on the reference height,which is usually set as the lowest level in the atmosphere in Earth system models,has not received much attention.Based on high-resolution large-eddy simulation(LES)data under unstable conditions,we find the setting of reference height is not trivial within the framework of current surface layer theory.With a reasonable prescription of aerodynamic roughness length(following the setting in LESs),reference heights near the top of the surface layer tend to provide the best estimate of surface fluxes,especially for the momentum flux.Furthermore,this conclusion for the sensible heat flux is insensitive to the ratio of roughness length for momentum versus heat.These results are robust,whether using the classical or revised surface layer theory.They provide a potential guide for setting the proper reference heights for Earth system modeling and can be further tested in the near future using observational data from land–atmosphere feedback observatories.展开更多
Graph Neural Networks(GNNs),as a deep learning framework specifically designed for graph-structured data,have achieved deep representation learning of graph data through message passing mechanisms and have become a co...Graph Neural Networks(GNNs),as a deep learning framework specifically designed for graph-structured data,have achieved deep representation learning of graph data through message passing mechanisms and have become a core technology in the field of graph analysis.However,current reviews on GNN models are mainly focused on smaller domains,and there is a lack of systematic reviews on the classification and applications of GNN models.This review systematically synthesizes the three canonical branches of GNN,Graph Convolutional Network(GCN),Graph Attention Network(GAT),and Graph Sampling Aggregation Network(GraphSAGE),and analyzes their integration pathways from both structural and feature perspectives.Drawing on representative studies,we identify three major integration patterns:cascaded fusion,where heterogeneous modules such as Convolutional Neural Network(CNN),Long Short-Term Memory(LSTM),and GraphSAGE are sequentially combined for hierarchical feature learning;parallel fusion,where multi-branch architectures jointly encode complementary graph features;and feature-level fusion,which employs concatenation,weighted summation,or attention-based gating to adaptively merge multi-source embeddings.Through these patterns,integrated GNNs achieve enhanced expressiveness,robustness,and scalability across domains including transportation,biomedicine,and cybersecurity.展开更多
Phase Two of the Integrative Monsoon Frontal Rainfall Experiment(IMFRE-II)was conducted over the middle and lower reaches of the Yangtze River during the period 16 June to 19 July 2020.This paper provides a brief over...Phase Two of the Integrative Monsoon Frontal Rainfall Experiment(IMFRE-II)was conducted over the middle and lower reaches of the Yangtze River during the period 16 June to 19 July 2020.This paper provides a brief overview of the IMFRE-II field campaign,including the multiple ground-based remote sensors,aircraft probes,and their corresponding measurements during the 2020 mei-yu period,as well as how to use these numerous datasets to answer scientific questions.The highlights of IMFRE-II are:(1)to the best of our knowledge,IMFRE-II is the first field campaign in China to use ground-based,airborne,and spaceborne platforms to conduct comprehensive observations over the middle and lower reaches of the Yangtze River;and(2)seven aircraft flights were successfully carried out,and the spectra of ice particles,cloud droplets,and raindrops at different altitudes were obtained.These in-situ measurements will provide a“cloud truth”to validate the ground-based and satellite-retrieved cloud and precipitation properties and quantitatively estimate their retrieval uncertainties.They are also crucial for the development of a warm(and/or cold)rain conceptual model in order to better understand the cloud-to-rain conversion and accretion processes in mei-yu precipitation events.Through an integrative analysis of ground-based,aircraft,and satellite observations and model simulations,we can significantly improve our cloud and precipitation retrieval algorithms,investigate the microphysical properties of cloud and precipitation,understand in-depth the formation and dissipation mechanisms of mei-yu frontal systems,and improve cloud microphysics parameterization schemes and model simulations.展开更多
Recent attention has been put into recurring slope lineae (RSL), after the discovery that water is present in them. It is assumed that RSL are due to flowing water. However, even though that might be the case, the gen...Recent attention has been put into recurring slope lineae (RSL), after the discovery that water is present in them. It is assumed that RSL are due to flowing water. However, even though that might be the case, the general characteristics of RSL as well as their seasonal and spatial distribution in Mars, and their occurrence within craters, suggest that RSL correspond to the weathering of frozen aquifers, which coincides with slope stability processes occurring in impact craters and scree slopes from Earth. In this study, we associated RSL with similar weathering processes occurring on impact craters and hydrogeological processes occurring on Earth (including ice, water, and wind erosion and natural aquifer recharge processes). We were able to create a conceptual model on how RSL develop, why are they found mostly in mid latitudes around craters, why are they present in more frequency in one side of crates in high latitudes, and why are there more RSL in the Martian southern hemisphere. Considering the whole hydrogeological processes occurring in craters that experience RSL, we were able to predict where large quantities of liquid water are most likely to be present in the red planet.展开更多
Artificial intelligence(AI)has evolved at an unprecedented pace in recent years.This rapid advancement includes algorithmic breakthroughs,cross-disciplinary integration,and diverse applications—driven by growing comp...Artificial intelligence(AI)has evolved at an unprecedented pace in recent years.This rapid advancement includes algorithmic breakthroughs,cross-disciplinary integration,and diverse applications—driven by growing computational power,massive datasets,and collaborative global research.This special issue of Emerging Artificial Intelligence Technologies and Applications was conceived to provide a platformfor cuttingedge AI research communication,developing novel methodologies,cross-domain applications,and critical advancements in addressing real-world challenges.Over the past months,we have witnessed a remarkable diversity of submissions,reflecting the global trend of AI innovation.Below,we synthesize the key insights from these works,highlighting their collective contribution to advancing AI’s theoretical frontiers and practical applications.展开更多
A DFN-DEC(discrete fracture network-distinct element code)method based on the MATLAB platform is developed to generate heterogeneous DFN.Subsequently,the effects of the spatial variability(the meanμand the standard d...A DFN-DEC(discrete fracture network-distinct element code)method based on the MATLAB platform is developed to generate heterogeneous DFN.Subsequently,the effects of the spatial variability(the meanμand the standard deviationσ)of the geometric properties(i.e.,the fracture dip D,the trace length T and the spacing S)of both the gently-dipping(denoted with 1)and the steeply-dipping(denoted with 2)fractures on the stability of granite slope are investigated.Results indicate that the proposed DFN-DEC method is robust,generating fracture networks that resemble reality.In addition,the spatial variability of fracture geometry,influencing the structure of granite slope,plays a significant role in slope stability.The mean stability of the slope decreases with the increase ofμ_(D_(1))(the mean of gently-dipping fracture dip),σ_(D_(2))(the mean of steeply-dipping fracture dip),μ_(T_(1))(the mean of gently-dipping fracture trace length),μ_(T_(2))(the mean of steeply-dipping fracture trace length),σ_(T_(1))(the standard deviation of gently-dipping fracture trace length),σ_(T_(2))(the standard deviation of steeply-dipping fracture trace length),and the decrease ofσ_(D_(1))(the standard deviation of gently-dipping fracture dip),μ_(D_(2))(the standard deviation of steeply-dipping fracture dip),μ_(S_(1))(the mean of gently-dipping fracture spacing)andμ_(S_(2))(the mean of steeply-dipping fracture spacing).Among them,μ_(T_(1)),μ_(D_(1))andμ_(S_(1))have the major impact.When the fracture spacing is large,the variability in the fracture geometry becomes less relevant to slope stability.When within some ranges of the fracture spacing,the spatial varying of dips can increase the slope stability by forming an interlaced structure.The results also show that the effects of the variability of trace length on slope stability depend on the variability of dip.These findings highlight the importance of spatial variability in the geometry of fractures to rock slope stability analysis.展开更多
The treatment of aerosols, clouds, radiation, and precipitation in climate models, in addition to their interactions and as- sociated feedbacks, has long been one of the largest sources of uncertainty in predicting an...The treatment of aerosols, clouds, radiation, and precipitation in climate models, in addition to their interactions and as- sociated feedbacks, has long been one of the largest sources of uncertainty in predicting any potential future climate changes. Although many improvements have been made in CMIP5, aerosols, clouds, radiation, and their feedbacks are still a problem in climate models, as concluded in IPCC AR5 and published papers. Many studies have shown that modeled aerosols, clouds, radiation, and precipitation agree with observations within a certain range on a global scale; however, large biases occur at the regional scale. Characterizing the effects of aerosols and clouds on energy and the hydrological cycle and understanding the interactions among aerosols, clouds, radiation, and precipitation, are critical for weather forecasting and climate models. Significant improvements are needed, which require advanced observations and modeling at a range of spatial and temporal scales.展开更多
A closed-cell marine stratocumulus case during the Aerosol and Cloud Experiments in the Eastern North Atlantic(ACE-ENA)aircraft field campaign is selected to examine the heterogeneities of cloud and drizzle microphysi...A closed-cell marine stratocumulus case during the Aerosol and Cloud Experiments in the Eastern North Atlantic(ACE-ENA)aircraft field campaign is selected to examine the heterogeneities of cloud and drizzle microphysical properties and the aerosol-cloud-precipitation interactions.The spatial and vertical variabilities of cloud and drizzle microphysics are found in two different sets of flight legs:Leg-1 and Leg-2,which are parallel and perpendicular to the cloud propagation,respectively.The cloud along Leg-2 was close to adiabatic,where cloud-droplet effective radius and liquid water content linearly increase from cloud base to cloud top with less drizzle.The cloud along Leg-1 was sub-adiabatic with lower clouddroplet number concentration and larger cloud-droplet effective,but higher drizzle droplet number concentration,larger drizzle droplet median diameter and drizzle liquid water content.The heavier drizzle frequency and intensity on Leg-1 were enhanced by the collision-coalescence processes within cloud due to strong turbulence.The sub-cloud precipitation rate on Leg-1 was significantly higher than that along Leg-2.As a result,the sub-cloud accumulation mode aerosols and CCN on Leg-1 were depleted,but the coarse model aerosols increased.This further leads to a counter-intuitive phenomenon that the CCN is less than cloud-droplet number concentration for Leg-1.The average CCN loss rates are −3.89 cm^(-3)h^(-1)and −0.77 cm^(-3)h^(-1) on Leg-1 and Leg-2,respectively.The cloud and drizzle heterogeneities inside the same stratocumulus can significantly alter the sub-cloud aerosols and CCN budget.Hence it should be treated with caution in the aircraft assessment of aerosol-cloud-precipitation interactions.展开更多
The Southern Ocean is covered by a large amount of clouds with high cloud albedo.However,as reported by previous climate model intercomparison projects,underestimated cloudiness and overestimated absorption of solar r...The Southern Ocean is covered by a large amount of clouds with high cloud albedo.However,as reported by previous climate model intercomparison projects,underestimated cloudiness and overestimated absorption of solar radiation(ASR)over the Southern Ocean lead to substantial biases in climate sensitivity.The present study revisits this long-standing issue and explores the uncertainty sources in the latest CMIP6 models.We employ 10-year satellite observations to evaluate cloud radiative effect(CRE)and cloud physical properties in five CMIP6 models that provide comprehensive output of cloud,radiation,and aerosol.The simulated longwave,shortwave,and net CRE at the top of atmosphere in CMIP6 are comparable with the CERES satellite observations.Total cloud fraction(CF)is also reasonably simulated in CMIP6,but the comparison of liquid cloud fraction(LCF)reveals marked biases in spatial pattern and seasonal variations.The discrepancies between the CMIP6 models and the MODIS satellite observations become even larger in other cloud macroand micro-physical properties,including liquid water path(LWP),cloud optical depth(COD),and cloud effective radius,as well as aerosol optical depth(AOD).However,the large underestimation of both LWP and cloud effective radius(regional means~20%and 11%,respectively)results in relatively smaller bias in COD,and the impacts of the biases in COD and LCF also cancel out with each other,leaving CRE and ASR reasonably predicted in CMIP6.An error estimation framework is employed,and the different signs of the sensitivity errors and biases from CF and LWP corroborate the notions that there are compensating errors in the modeled shortwave CRE.Further correlation analyses of the geospatial patterns reveal that CF is the most relevant factor in determining CRE in observations,while the modeled CRE is too sensitive to LWP and COD.The relationships between cloud effective radius,LWP,and COD are also analyzed to explore the possible uncertainty sources in different models.Our study calls for more rigorous calibration of detailed cloud physical properties for future climate model development and climate projection.展开更多
Aerosol-cloud-radiation interactions represent one of the largest uncertainties in the current climate assessment. Much of the complexity arises from the non-monotonic responses of clouds, precipitation and radiative ...Aerosol-cloud-radiation interactions represent one of the largest uncertainties in the current climate assessment. Much of the complexity arises from the non-monotonic responses of clouds, precipitation and radiative fluxes to aerosol perturbations under various meteorological conditions. In this study, an aerosol-aware WRF model is used to investigate the microphysical and radiative effects of aerosols in three weather systems during the March 2000 Cloud Intensive Observational Period campaign at the US Southern Great Plains. Three simulated cloud ensembles include a low-pressure deep convective cloud system, a collection of less-precipitating stratus and shallow cumulus, and a cold frontal passage. The WRF simulations are evaluated by several ground-based measurements. The microphysical properties of cloud hydrometeors, such as their mass and number concentrations, generally show monotonic trends as a function of cloud condensation nuclei concentrations. Aerosol radiative effects do not influence the trends of cloud microphysics, except for the stratus and shallow cumulus cases where aerosol semi-direct effects are identified. The precipitation changes by aerosols vary with the cloud types and their evolving stages, with a prominent aerosol invigoration effect and associated enhanced precipitation from the convective sources. The simulated aerosol direct effect suppresses precipitation in all three cases but does not overturn the aerosol indirect effect. Cloud fraction exhibits much smaller sensitivity （typically less than 2%） to aerosol perturbations, and the responses vary with aerosol concentrations and cloud regimes. The surface shortwave radiation shows a monotonic decrease by increasing aerosols, while the magnitude of the decrease depends on the cloud type.展开更多
Restricted by the development of the transient flow and solute reactive transport models for unsaturated soil, empirical functions have been used previously to calculate the mass of dissolved or precipitated salt when...Restricted by the development of the transient flow and solute reactive transport models for unsaturated soil, empirical functions have been used previously to calculate the mass of dissolved or precipitated salt when they have to be taken into account. Besides, the solute reactive transport process has often been inferred based on measurements that cost lots of time and manpower. HP1 model coupled with PHREEQC provides a suitable tool to improve the estimation of salt distribution during evaporation in saline soil, where the salt dissolution and precipitation cannot be ignored. In this study, we compare the performance of a standard solute transport(SST) model and the HP1 model to examine the improvement of salt distribution estimation. Model results are compared with experimental data sets from four field lysimeters. These columns were exposed to Na Cl solution with different concentrations(3, 30, 100, and 250 g/L) and were undergoing the same strong evaporation boundary condition. The pre-existing Ca SO_(4), Na Cl and Na2SO_(4)loads were 1.15, 0.47 and 0.23 g/(100 g of soil), respectively. Simulation results show that HP1 ameliorates the overestimation of salt content by SST in deeper soil due to the absence of dissolution of pre-existing soluble salts, and prevents the concentration of the solute from exceeding the solubilities which would occur in SST-result. Additionally, HP1-predicted results can help trace the transport process of each solute. Based on the results, we strongly suggest that the management of fields sensitive to salt content should make use of a coupled flow and chemical reaction model.展开更多
Drought is one of the extreme events that can be caused by internal climate variability （ICV） and external forcing （EF）. Here, the authors investigate the relative contributions of ICY and EF to meteorological dro...Drought is one of the extreme events that can be caused by internal climate variability （ICV） and external forcing （EF）. Here, the authors investigate the relative contributions of ICY and EF to meteorological drought changes in China using 40 members from the Community Earth System Model Large Ensemble （CESE_LE） project for historical simulations （in response to greenhouse gases and other EF） and future simulations under the RCP8.5 scenario. The authors use the Standardized Precipitation Index （SPI） to represent meteorological drought, and then define and analyze four drought parameters （frequency, severity, duration, and maximum duration） over eight regions of China. For historical periods, the ICV plays a dominant role in drought variation, while with global warming under the RCP8.5 scenario the EF becomes the prominent factor for drought characteristics. With the global warming signal, the effect of ICV varies with the drought parameters. This study suggests that the ICV should be taken into account when climate model simulations are used to investigate drought--in particular, for historical periods.展开更多
Throughout vast areas of Asia,the summer of 2020 was extraordinarily wet.After an exceptionally wet May in Northeast India and Bangladesh,excessive rainfall hit at least 10 provinces in central and southern China in J...Throughout vast areas of Asia,the summer of 2020 was extraordinarily wet.After an exceptionally wet May in Northeast India and Bangladesh,excessive rainfall hit at least 10 provinces in central and southern China in June and July,causing extensive flooding in many rural and urban locations.Long standing rainfall,lake and river level records were consequently broken in several parts of the region with the Yangtze-Huaihe river valleys,particularly badly impacted,with consequential economic losses.Floods and landslides also affected parts of Japan with at least one location in Kumamoto province even experiencing a record-breaking 1000 mm of rainfall in just 3 days in early July.The 2020 wet season in South Korea was also exceptionally long,lasting 54 days,compared to their more usual 32.展开更多
Recent Super-Resolution(SR)algorithms often suffer from excessive model complexity,high computational costs,and limited flexibility across varying image scales.To address these challenges,we propose DDNet,a dynamic an...Recent Super-Resolution(SR)algorithms often suffer from excessive model complexity,high computational costs,and limited flexibility across varying image scales.To address these challenges,we propose DDNet,a dynamic and lightweight SR framework designed for arbitrary scaling factors.DDNet integrates a residual learning structure with an Adaptively fusion Feature Block(AFB)and a scale-aware upsampling module,effectively reducing parameter overhead while preserving reconstruction quality.Additionally,we introduce DDNetGAN,an enhanced variant that leverages a relativistic Generative Adversarial Network(GAN)to further improve texture realism.To validate the proposed models,we conduct extensive training using the DIV2K and Flickr2K datasets and evaluate performance across standard benchmarks including Set5,Set14,Urban100,Manga109,and BSD100.Our experiments cover both symmetric and asymmetric upscaling factors and incorporate ablation studies to assess key components.Results show that DDNet and DDNetGAN achieve competitive performance compared with mainstream SR algorithms,demonstrating a strong balance between accuracy,efficiency,and flexibility.These findings highlight the potential of our approach for practical real-world super-resolution applications.展开更多
Satellite Precipitation Products(SPPs) face challenges in detecting Extreme Precipitation Events(EPEs). Hence, the primary objective of this research is to introduce a novel framework termed Machine-Learning Clusterin...Satellite Precipitation Products(SPPs) face challenges in detecting Extreme Precipitation Events(EPEs). Hence, the primary objective of this research is to introduce a novel framework termed Machine-Learning Clustering-Merging Algorithms(ML-CMAs) to evaluate EPEs using SPPs and Auxiliary Data(AD). Daily precipitation measurements were utilized for training and evaluating EPE estimates over Iran, which is comprised of arid and semi-arid regions. Statistical analysis and evaluation of five SPPs demonstrated that during EPE occurrences, all products face challenges in precipitation estimation, and using these products individually is not recommended. Among the SPPs, Multi-Source Weighted-Ensemble Precipitation(MSWEP) performed best for heavy(>20 mm d–1) and extreme(>40 mm d–1)precipitation events, followed by Global Satellite Mapping of Precipitation(GSMa P), Integrated Multi-Satellite Retrievals for Global Precipitation Measurement(IMERG), Climate Prediction Center morphing technique(CMORPH), and Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks Dynamic Infrared-Rain Rate(PERSIANN-PDIR). The findings indicate that all proposed methods based on ML-CMAs could estimate precipitation rates more accurately than SPPs and improve statistical indices. The seasonal assessment and spatial analysis of statistical metrics of the overall daily precipitation results for all periods and climates revealed that all methods based on ML-CMAs performed well in all seasons and at nearly all measurement stations. Using unsupervised K-means++ classification for clustering EPEs and Deep Neural Network(DNN) and Convolutional Neural Network(CNN) methods for merging the MLCMAs reduced the error rate of SPPs in EPE estimation by approximately 50%. Therefore, incorporating ML-CMAs along with PWV as AD can significantly improve the performance of SPPs in evaluating EPEs over the study region.展开更多
Eastward-moving cloud clusters from the Tibetan Plateau(TP)often trigger heavy rainfall events in the Yangtze River basin in summer.Forecasting these events in an operational environment remains a challenging task.Her...Eastward-moving cloud clusters from the Tibetan Plateau(TP)often trigger heavy rainfall events in the Yangtze River basin in summer.Forecasting these events in an operational environment remains a challenging task.Here,dynamical diagnosis and a Lagrangian trajectory model are used to analyze the background atmospheric circulation,maintenance mechanism,and moisture transport of two Meiyu front rainstorms(MYFR)during 30 June-2 July 2016 and 17-19 June 2018 associated with eastward-moving cloud clusters from the TP.It is shown that in both cases heavy rainfall is characterized by semi-continuous rainbelts extending from the eastern TP to the Yangtze River valleys with eastward-spreading convective clouds weakening and strengthening alternately from the eastern TP to downstream regions.Following the track of positive water vapor advection,centers of positive vorticity propagate downstream through the Sichuan basin.The baroclinic thermodynamic–dynamical interaction and the barotropic nonequilibrium force work against each other in the development of the MYFR.Specifically,during the early stage of precipitation development,the barotropic non-equilibrium force dominates,while during the period of heavy precipitation the baroclinic thermodynamic-dynamical interaction dominates.The convergence associated with the baroclinic thermodynamic-dynamical interaction guarantees the persistence of heavy precipitation.Compared to the climate mean state(1988-2018),both MYFR events associated with eastward-moving cloud clusters from the eastern TP are characterized by increased moisture transport from the southwest.One of the major paths of moisture transport in both cases is along the south side of the TP,directly connected to the eastward movement of cloud clusters.展开更多
The mei-yu season,typically occurring from mid-June to mid-July,on average,contributes to 32%of the annual precipitation over the Yangtze-Huai River Valley(YHRV)and represents one of the three heavy-rainfall periods i...The mei-yu season,typically occurring from mid-June to mid-July,on average,contributes to 32%of the annual precipitation over the Yangtze-Huai River Valley(YHRV)and represents one of the three heavy-rainfall periods in China.Here,we briefly review the large-scale background,synoptic pattern,moisture transport,and cloud and precipitation characteristics of the mei-yu frontal systems in the context of the ongoing Integrative Monsoon Frontal Rainfall Experiment(IMFRE)field campaign.Phase one of the campaign,IMFRE-I,was conducted from 10 June to 10 July 2018 in the middle reaches of the YHRV.Led by the Wuhan Institute of Heavy Rain(IHR)with primary support from the National Natural Science Foundation of China,IMFRE-I maximizes the use of our observational capacity enabled by a suite of ground-based and remote sensing instruments,most notably the IHR Mesoscale Heavy Rainfall Observing System(MHROS),including different wavelengths of radars,microwave radiometers,and disdrometers.The KA350(Shanxi King-Air)aircraft participating in the campaign is equipped with Ka-band cloud radar and different probes.The comprehensive datasets from both the MHROS and aircraft instruments are combined with available satellite observations and model simulations to answer the three scientific questions of IMFRE-I.Some highlights from a previously published special issue are included in this review,and we also briefly introduce the IMFRE-II field campaign,conducted during June-July 2020,where the focus was on the spatiotemporal evolutions of the mei-yu frontal systems over the middle and lower reaches of the YHRV.展开更多
基金supported by the Natural Science Foundation of China(Grant Nos.42088101 and 42375163)the Guangdong Major Project of Basic and Applied Basic Research(Grant No.2021B0301030007)the specific research fund of The Innovation Platform for Academicians of Hainan Province(Grant No.YSPTZX202143)。
文摘In Earth system modeling,the land surface is coupled with the atmosphere through surface turbulent fluxes.These fluxes are computed using mean meteorological variables between the surface and a reference height in the atmosphere.However,the dependence of flux computation on the reference height,which is usually set as the lowest level in the atmosphere in Earth system models,has not received much attention.Based on high-resolution large-eddy simulation(LES)data under unstable conditions,we find the setting of reference height is not trivial within the framework of current surface layer theory.With a reasonable prescription of aerodynamic roughness length(following the setting in LESs),reference heights near the top of the surface layer tend to provide the best estimate of surface fluxes,especially for the momentum flux.Furthermore,this conclusion for the sensible heat flux is insensitive to the ratio of roughness length for momentum versus heat.These results are robust,whether using the classical or revised surface layer theory.They provide a potential guide for setting the proper reference heights for Earth system modeling and can be further tested in the near future using observational data from land–atmosphere feedback observatories.
基金funded by Guangzhou Huashang University(2024HSZD01,HS2023JYSZH01).
文摘Graph Neural Networks(GNNs),as a deep learning framework specifically designed for graph-structured data,have achieved deep representation learning of graph data through message passing mechanisms and have become a core technology in the field of graph analysis.However,current reviews on GNN models are mainly focused on smaller domains,and there is a lack of systematic reviews on the classification and applications of GNN models.This review systematically synthesizes the three canonical branches of GNN,Graph Convolutional Network(GCN),Graph Attention Network(GAT),and Graph Sampling Aggregation Network(GraphSAGE),and analyzes their integration pathways from both structural and feature perspectives.Drawing on representative studies,we identify three major integration patterns:cascaded fusion,where heterogeneous modules such as Convolutional Neural Network(CNN),Long Short-Term Memory(LSTM),and GraphSAGE are sequentially combined for hierarchical feature learning;parallel fusion,where multi-branch architectures jointly encode complementary graph features;and feature-level fusion,which employs concatenation,weighted summation,or attention-based gating to adaptively merge multi-source embeddings.Through these patterns,integrated GNNs achieve enhanced expressiveness,robustness,and scalability across domains including transportation,biomedicine,and cybersecurity.
基金The IMFRE-II field campaign was primarily supported by the National Natural Science Foundation of China(Grant Nos.41620104009 and 91637211)the Key Program for International S&T Cooperation Projects of China(Grant No.2016YFE0109400)the National Key R&D Program of China(Grant No.2018YFC1507200).
文摘Phase Two of the Integrative Monsoon Frontal Rainfall Experiment(IMFRE-II)was conducted over the middle and lower reaches of the Yangtze River during the period 16 June to 19 July 2020.This paper provides a brief overview of the IMFRE-II field campaign,including the multiple ground-based remote sensors,aircraft probes,and their corresponding measurements during the 2020 mei-yu period,as well as how to use these numerous datasets to answer scientific questions.The highlights of IMFRE-II are:(1)to the best of our knowledge,IMFRE-II is the first field campaign in China to use ground-based,airborne,and spaceborne platforms to conduct comprehensive observations over the middle and lower reaches of the Yangtze River;and(2)seven aircraft flights were successfully carried out,and the spectra of ice particles,cloud droplets,and raindrops at different altitudes were obtained.These in-situ measurements will provide a“cloud truth”to validate the ground-based and satellite-retrieved cloud and precipitation properties and quantitatively estimate their retrieval uncertainties.They are also crucial for the development of a warm(and/or cold)rain conceptual model in order to better understand the cloud-to-rain conversion and accretion processes in mei-yu precipitation events.Through an integrative analysis of ground-based,aircraft,and satellite observations and model simulations,we can significantly improve our cloud and precipitation retrieval algorithms,investigate the microphysical properties of cloud and precipitation,understand in-depth the formation and dissipation mechanisms of mei-yu frontal systems,and improve cloud microphysics parameterization schemes and model simulations.
文摘Recent attention has been put into recurring slope lineae (RSL), after the discovery that water is present in them. It is assumed that RSL are due to flowing water. However, even though that might be the case, the general characteristics of RSL as well as their seasonal and spatial distribution in Mars, and their occurrence within craters, suggest that RSL correspond to the weathering of frozen aquifers, which coincides with slope stability processes occurring in impact craters and scree slopes from Earth. In this study, we associated RSL with similar weathering processes occurring on impact craters and hydrogeological processes occurring on Earth (including ice, water, and wind erosion and natural aquifer recharge processes). We were able to create a conceptual model on how RSL develop, why are they found mostly in mid latitudes around craters, why are they present in more frequency in one side of crates in high latitudes, and why are there more RSL in the Martian southern hemisphere. Considering the whole hydrogeological processes occurring in craters that experience RSL, we were able to predict where large quantities of liquid water are most likely to be present in the red planet.
文摘Artificial intelligence(AI)has evolved at an unprecedented pace in recent years.This rapid advancement includes algorithmic breakthroughs,cross-disciplinary integration,and diverse applications—driven by growing computational power,massive datasets,and collaborative global research.This special issue of Emerging Artificial Intelligence Technologies and Applications was conceived to provide a platformfor cuttingedge AI research communication,developing novel methodologies,cross-domain applications,and critical advancements in addressing real-world challenges.Over the past months,we have witnessed a remarkable diversity of submissions,reflecting the global trend of AI innovation.Below,we synthesize the key insights from these works,highlighting their collective contribution to advancing AI’s theoretical frontiers and practical applications.
基金supported by the National Natural Science Foundation of China(Nos.41807264,41972289)the Engineering Research Center of Rock-Soil Drilling&Excavation and Protection,Ministry of Education(No.202102)+3 种基金the Key Laboratory of Geological Hazards on Three Gorges Reservoir Area(China Three Gorges University),Ministry of Education(No.2020KDZ01)the Fundamental Research Funds for the Central Universities,China University of Geosciences(Wuhan)(Nos.CUG170686,CUGQY1932)the China Scholarship Council(No.201406410032)the Science and Technology Research Project of Education Department of Hubei Province(Nos.B2019452,B2024509)。
文摘A DFN-DEC(discrete fracture network-distinct element code)method based on the MATLAB platform is developed to generate heterogeneous DFN.Subsequently,the effects of the spatial variability(the meanμand the standard deviationσ)of the geometric properties(i.e.,the fracture dip D,the trace length T and the spacing S)of both the gently-dipping(denoted with 1)and the steeply-dipping(denoted with 2)fractures on the stability of granite slope are investigated.Results indicate that the proposed DFN-DEC method is robust,generating fracture networks that resemble reality.In addition,the spatial variability of fracture geometry,influencing the structure of granite slope,plays a significant role in slope stability.The mean stability of the slope decreases with the increase ofμ_(D_(1))(the mean of gently-dipping fracture dip),σ_(D_(2))(the mean of steeply-dipping fracture dip),μ_(T_(1))(the mean of gently-dipping fracture trace length),μ_(T_(2))(the mean of steeply-dipping fracture trace length),σ_(T_(1))(the standard deviation of gently-dipping fracture trace length),σ_(T_(2))(the standard deviation of steeply-dipping fracture trace length),and the decrease ofσ_(D_(1))(the standard deviation of gently-dipping fracture dip),μ_(D_(2))(the standard deviation of steeply-dipping fracture dip),μ_(S_(1))(the mean of gently-dipping fracture spacing)andμ_(S_(2))(the mean of steeply-dipping fracture spacing).Among them,μ_(T_(1)),μ_(D_(1))andμ_(S_(1))have the major impact.When the fracture spacing is large,the variability in the fracture geometry becomes less relevant to slope stability.When within some ranges of the fracture spacing,the spatial varying of dips can increase the slope stability by forming an interlaced structure.The results also show that the effects of the variability of trace length on slope stability depend on the variability of dip.These findings highlight the importance of spatial variability in the geometry of fractures to rock slope stability analysis.
基金primarily supported by the Major International (Regional) Joint Research Project of the National Science Foundation of China (NSFC) (Grant No. 41620104009) at the Institute of Heavy Rain
文摘The treatment of aerosols, clouds, radiation, and precipitation in climate models, in addition to their interactions and as- sociated feedbacks, has long been one of the largest sources of uncertainty in predicting any potential future climate changes. Although many improvements have been made in CMIP5, aerosols, clouds, radiation, and their feedbacks are still a problem in climate models, as concluded in IPCC AR5 and published papers. Many studies have shown that modeled aerosols, clouds, radiation, and precipitation agree with observations within a certain range on a global scale; however, large biases occur at the regional scale. Characterizing the effects of aerosols and clouds on energy and the hydrological cycle and understanding the interactions among aerosols, clouds, radiation, and precipitation, are critical for weather forecasting and climate models. Significant improvements are needed, which require advanced observations and modeling at a range of spatial and temporal scales.
基金supported by the NSF grants AGS-2031750 and AGS-2031751supported as part of the “Enabling Aerosol-cloud interactions at GLobal convection-permitting scal ES (EAGLES)” project (74358),funded by the U.S. Department of Energy, Office of Science, Office of Biological and Environmental Research, Earth System Modeling program with the subcontract to the University of Arizona
文摘A closed-cell marine stratocumulus case during the Aerosol and Cloud Experiments in the Eastern North Atlantic(ACE-ENA)aircraft field campaign is selected to examine the heterogeneities of cloud and drizzle microphysical properties and the aerosol-cloud-precipitation interactions.The spatial and vertical variabilities of cloud and drizzle microphysics are found in two different sets of flight legs:Leg-1 and Leg-2,which are parallel and perpendicular to the cloud propagation,respectively.The cloud along Leg-2 was close to adiabatic,where cloud-droplet effective radius and liquid water content linearly increase from cloud base to cloud top with less drizzle.The cloud along Leg-1 was sub-adiabatic with lower clouddroplet number concentration and larger cloud-droplet effective,but higher drizzle droplet number concentration,larger drizzle droplet median diameter and drizzle liquid water content.The heavier drizzle frequency and intensity on Leg-1 were enhanced by the collision-coalescence processes within cloud due to strong turbulence.The sub-cloud precipitation rate on Leg-1 was significantly higher than that along Leg-2.As a result,the sub-cloud accumulation mode aerosols and CCN on Leg-1 were depleted,but the coarse model aerosols increased.This further leads to a counter-intuitive phenomenon that the CCN is less than cloud-droplet number concentration for Leg-1.The average CCN loss rates are −3.89 cm^(-3)h^(-1)and −0.77 cm^(-3)h^(-1) on Leg-1 and Leg-2,respectively.The cloud and drizzle heterogeneities inside the same stratocumulus can significantly alter the sub-cloud aerosols and CCN budget.Hence it should be treated with caution in the aircraft assessment of aerosol-cloud-precipitation interactions.
基金supported by the National Science Foundation grants(Grant Nos.AGS-1700727/1700728,2031751/2031750)supported by the National Natural Science Foundation of China.(Grant No.41925022).
文摘The Southern Ocean is covered by a large amount of clouds with high cloud albedo.However,as reported by previous climate model intercomparison projects,underestimated cloudiness and overestimated absorption of solar radiation(ASR)over the Southern Ocean lead to substantial biases in climate sensitivity.The present study revisits this long-standing issue and explores the uncertainty sources in the latest CMIP6 models.We employ 10-year satellite observations to evaluate cloud radiative effect(CRE)and cloud physical properties in five CMIP6 models that provide comprehensive output of cloud,radiation,and aerosol.The simulated longwave,shortwave,and net CRE at the top of atmosphere in CMIP6 are comparable with the CERES satellite observations.Total cloud fraction(CF)is also reasonably simulated in CMIP6,but the comparison of liquid cloud fraction(LCF)reveals marked biases in spatial pattern and seasonal variations.The discrepancies between the CMIP6 models and the MODIS satellite observations become even larger in other cloud macroand micro-physical properties,including liquid water path(LWP),cloud optical depth(COD),and cloud effective radius,as well as aerosol optical depth(AOD).However,the large underestimation of both LWP and cloud effective radius(regional means~20%and 11%,respectively)results in relatively smaller bias in COD,and the impacts of the biases in COD and LCF also cancel out with each other,leaving CRE and ASR reasonably predicted in CMIP6.An error estimation framework is employed,and the different signs of the sensitivity errors and biases from CF and LWP corroborate the notions that there are compensating errors in the modeled shortwave CRE.Further correlation analyses of the geospatial patterns reveal that CF is the most relevant factor in determining CRE in observations,while the modeled CRE is too sensitive to LWP and COD.The relationships between cloud effective radius,LWP,and COD are also analyzed to explore the possible uncertainty sources in different models.Our study calls for more rigorous calibration of detailed cloud physical properties for future climate model development and climate projection.
基金support provided by NASA ROSES14-ACMAPNSF (Award No. 1700727)+1 种基金supported by the US DOE ASR programsupport of the Jet Propulsion Laboratory, California Institute of Technology, under contract with NASA
文摘Aerosol-cloud-radiation interactions represent one of the largest uncertainties in the current climate assessment. Much of the complexity arises from the non-monotonic responses of clouds, precipitation and radiative fluxes to aerosol perturbations under various meteorological conditions. In this study, an aerosol-aware WRF model is used to investigate the microphysical and radiative effects of aerosols in three weather systems during the March 2000 Cloud Intensive Observational Period campaign at the US Southern Great Plains. Three simulated cloud ensembles include a low-pressure deep convective cloud system, a collection of less-precipitating stratus and shallow cumulus, and a cold frontal passage. The WRF simulations are evaluated by several ground-based measurements. The microphysical properties of cloud hydrometeors, such as their mass and number concentrations, generally show monotonic trends as a function of cloud condensation nuclei concentrations. Aerosol radiative effects do not influence the trends of cloud microphysics, except for the stratus and shallow cumulus cases where aerosol semi-direct effects are identified. The precipitation changes by aerosols vary with the cloud types and their evolving stages, with a prominent aerosol invigoration effect and associated enhanced precipitation from the convective sources. The simulated aerosol direct effect suppresses precipitation in all three cases but does not overturn the aerosol indirect effect. Cloud fraction exhibits much smaller sensitivity （typically less than 2%） to aerosol perturbations, and the responses vary with aerosol concentrations and cloud regimes. The surface shortwave radiation shows a monotonic decrease by increasing aerosols, while the magnitude of the decrease depends on the cloud type.
基金supported by the National Natural Science Foundation of China (Nos.41572224,U1403282,51709232)the Fundamental Research Funds for the Central Universities,China University of Geosciences (Wuhan) (No.CUG170103)。
文摘Restricted by the development of the transient flow and solute reactive transport models for unsaturated soil, empirical functions have been used previously to calculate the mass of dissolved or precipitated salt when they have to be taken into account. Besides, the solute reactive transport process has often been inferred based on measurements that cost lots of time and manpower. HP1 model coupled with PHREEQC provides a suitable tool to improve the estimation of salt distribution during evaporation in saline soil, where the salt dissolution and precipitation cannot be ignored. In this study, we compare the performance of a standard solute transport(SST) model and the HP1 model to examine the improvement of salt distribution estimation. Model results are compared with experimental data sets from four field lysimeters. These columns were exposed to Na Cl solution with different concentrations(3, 30, 100, and 250 g/L) and were undergoing the same strong evaporation boundary condition. The pre-existing Ca SO_(4), Na Cl and Na2SO_(4)loads were 1.15, 0.47 and 0.23 g/(100 g of soil), respectively. Simulation results show that HP1 ameliorates the overestimation of salt content by SST in deeper soil due to the absence of dissolution of pre-existing soluble salts, and prevents the concentration of the solute from exceeding the solubilities which would occur in SST-result. Additionally, HP1-predicted results can help trace the transport process of each solute. Based on the results, we strongly suggest that the management of fields sensitive to salt content should make use of a coupled flow and chemical reaction model.
基金supported by the Key Project of the Ministry of Science and Technology of China[grant number2016YFA0602401]the National Natural Science Foundation of China[grant number 41275110]supported by the National Science Foundation[grant number AGS-0944101]
文摘Drought is one of the extreme events that can be caused by internal climate variability （ICV） and external forcing （EF）. Here, the authors investigate the relative contributions of ICY and EF to meteorological drought changes in China using 40 members from the Community Earth System Model Large Ensemble （CESE_LE） project for historical simulations （in response to greenhouse gases and other EF） and future simulations under the RCP8.5 scenario. The authors use the Standardized Precipitation Index （SPI） to represent meteorological drought, and then define and analyze four drought parameters （frequency, severity, duration, and maximum duration） over eight regions of China. For historical periods, the ICV plays a dominant role in drought variation, while with global warming under the RCP8.5 scenario the EF becomes the prominent factor for drought characteristics. With the global warming signal, the effect of ICV varies with the drought parameters. This study suggests that the ICV should be taken into account when climate model simulations are used to investigate drought--in particular, for historical periods.
基金Robin CLARK was supported by the UK-China Research&Innovation Partnership Fund through the Met Office Climate Science for Service Partnership(CSSP)China as part of the Newton Fund.Chang-Hoi HO was supported by Korea Meteorological Administration Research and Development Program(Grant No.KMI2020-00610)Tetsuya TAKEMI was supported,on this topic,by the Environment Research and Technology Development Fund(ERTDF)JPMEERF20192005 of the Environmental Restoration and Conservation Agency(ERCA)of Japan.
文摘Throughout vast areas of Asia,the summer of 2020 was extraordinarily wet.After an exceptionally wet May in Northeast India and Bangladesh,excessive rainfall hit at least 10 provinces in central and southern China in June and July,causing extensive flooding in many rural and urban locations.Long standing rainfall,lake and river level records were consequently broken in several parts of the region with the Yangtze-Huaihe river valleys,particularly badly impacted,with consequential economic losses.Floods and landslides also affected parts of Japan with at least one location in Kumamoto province even experiencing a record-breaking 1000 mm of rainfall in just 3 days in early July.The 2020 wet season in South Korea was also exceptionally long,lasting 54 days,compared to their more usual 32.
基金supported by Sichuan Science and Technology Program[2023YFSY0026,2023YFH0004].
文摘Recent Super-Resolution(SR)algorithms often suffer from excessive model complexity,high computational costs,and limited flexibility across varying image scales.To address these challenges,we propose DDNet,a dynamic and lightweight SR framework designed for arbitrary scaling factors.DDNet integrates a residual learning structure with an Adaptively fusion Feature Block(AFB)and a scale-aware upsampling module,effectively reducing parameter overhead while preserving reconstruction quality.Additionally,we introduce DDNetGAN,an enhanced variant that leverages a relativistic Generative Adversarial Network(GAN)to further improve texture realism.To validate the proposed models,we conduct extensive training using the DIV2K and Flickr2K datasets and evaluate performance across standard benchmarks including Set5,Set14,Urban100,Manga109,and BSD100.Our experiments cover both symmetric and asymmetric upscaling factors and incorporate ablation studies to assess key components.Results show that DDNet and DDNetGAN achieve competitive performance compared with mainstream SR algorithms,demonstrating a strong balance between accuracy,efficiency,and flexibility.These findings highlight the potential of our approach for practical real-world super-resolution applications.
文摘Satellite Precipitation Products(SPPs) face challenges in detecting Extreme Precipitation Events(EPEs). Hence, the primary objective of this research is to introduce a novel framework termed Machine-Learning Clustering-Merging Algorithms(ML-CMAs) to evaluate EPEs using SPPs and Auxiliary Data(AD). Daily precipitation measurements were utilized for training and evaluating EPE estimates over Iran, which is comprised of arid and semi-arid regions. Statistical analysis and evaluation of five SPPs demonstrated that during EPE occurrences, all products face challenges in precipitation estimation, and using these products individually is not recommended. Among the SPPs, Multi-Source Weighted-Ensemble Precipitation(MSWEP) performed best for heavy(>20 mm d–1) and extreme(>40 mm d–1)precipitation events, followed by Global Satellite Mapping of Precipitation(GSMa P), Integrated Multi-Satellite Retrievals for Global Precipitation Measurement(IMERG), Climate Prediction Center morphing technique(CMORPH), and Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks Dynamic Infrared-Rain Rate(PERSIANN-PDIR). The findings indicate that all proposed methods based on ML-CMAs could estimate precipitation rates more accurately than SPPs and improve statistical indices. The seasonal assessment and spatial analysis of statistical metrics of the overall daily precipitation results for all periods and climates revealed that all methods based on ML-CMAs performed well in all seasons and at nearly all measurement stations. Using unsupervised K-means++ classification for clustering EPEs and Deep Neural Network(DNN) and Convolutional Neural Network(CNN) methods for merging the MLCMAs reduced the error rate of SPPs in EPE estimation by approximately 50%. Therefore, incorporating ML-CMAs along with PWV as AD can significantly improve the performance of SPPs in evaluating EPEs over the study region.
基金Supported by the National Natural Science Foundation of China(41620104009 and 41975058)Science and Technology Funds of Hubei Meteorological Bureau(2022Y25 and 2022Z02)+3 种基金Joint Open Project of Key Laboratory of Meteorological Disaster,Ministry of Education&Collaborative Innovation Center on Forecast and Evaluation of Meteorological DisastersNanjing University of Information Science&Technology(KLME202106)in part supported by the U.S.National Science Foundation(AGS-2032532)NOAA(NA20OAR4310380)
文摘Eastward-moving cloud clusters from the Tibetan Plateau(TP)often trigger heavy rainfall events in the Yangtze River basin in summer.Forecasting these events in an operational environment remains a challenging task.Here,dynamical diagnosis and a Lagrangian trajectory model are used to analyze the background atmospheric circulation,maintenance mechanism,and moisture transport of two Meiyu front rainstorms(MYFR)during 30 June-2 July 2016 and 17-19 June 2018 associated with eastward-moving cloud clusters from the TP.It is shown that in both cases heavy rainfall is characterized by semi-continuous rainbelts extending from the eastern TP to the Yangtze River valleys with eastward-spreading convective clouds weakening and strengthening alternately from the eastern TP to downstream regions.Following the track of positive water vapor advection,centers of positive vorticity propagate downstream through the Sichuan basin.The baroclinic thermodynamic–dynamical interaction and the barotropic nonequilibrium force work against each other in the development of the MYFR.Specifically,during the early stage of precipitation development,the barotropic non-equilibrium force dominates,while during the period of heavy precipitation the baroclinic thermodynamic-dynamical interaction dominates.The convergence associated with the baroclinic thermodynamic-dynamical interaction guarantees the persistence of heavy precipitation.Compared to the climate mean state(1988-2018),both MYFR events associated with eastward-moving cloud clusters from the eastern TP are characterized by increased moisture transport from the southwest.One of the major paths of moisture transport in both cases is along the south side of the TP,directly connected to the eastward movement of cloud clusters.
基金The datasets were provided by the Mesoscale Heavy Rainfall Observing System(MHROS)of the Wuhan Institute of Heave Rain(IHR),China Meteorological AdministrationThe IMFRE field campaign is primarily supported by the National Natural Science Foundation of China(Grant Nos.41620104009 and 91637211).
文摘The mei-yu season,typically occurring from mid-June to mid-July,on average,contributes to 32%of the annual precipitation over the Yangtze-Huai River Valley(YHRV)and represents one of the three heavy-rainfall periods in China.Here,we briefly review the large-scale background,synoptic pattern,moisture transport,and cloud and precipitation characteristics of the mei-yu frontal systems in the context of the ongoing Integrative Monsoon Frontal Rainfall Experiment(IMFRE)field campaign.Phase one of the campaign,IMFRE-I,was conducted from 10 June to 10 July 2018 in the middle reaches of the YHRV.Led by the Wuhan Institute of Heavy Rain(IHR)with primary support from the National Natural Science Foundation of China,IMFRE-I maximizes the use of our observational capacity enabled by a suite of ground-based and remote sensing instruments,most notably the IHR Mesoscale Heavy Rainfall Observing System(MHROS),including different wavelengths of radars,microwave radiometers,and disdrometers.The KA350(Shanxi King-Air)aircraft participating in the campaign is equipped with Ka-band cloud radar and different probes.The comprehensive datasets from both the MHROS and aircraft instruments are combined with available satellite observations and model simulations to answer the three scientific questions of IMFRE-I.Some highlights from a previously published special issue are included in this review,and we also briefly introduce the IMFRE-II field campaign,conducted during June-July 2020,where the focus was on the spatiotemporal evolutions of the mei-yu frontal systems over the middle and lower reaches of the YHRV.
