Flow Patterns Research Articles

How to consider groundwater flow systems in the Earth's Critical Zone? – Demonstration in the Central Pannonian Basin, Hungary

Study regionCentral Pannonian Basin, Hungary. Study focusCritical Zone (CZ) Science generally focuses on the soil and weathered bedrock in a few or tens of meters depth, thus influence of deeper groundwater on the CZ is understudied. Here we aim to introduce a hydrogeological methodology that can separate normal and abnormal pressure regimes and determine the groundwater flow pattern to characterize the connection of different groundwater flow systems to the CZ. Basin-scale evaluation of about 5500 measured hydraulic data were carried out by p(z) and h(z) profiles, tomographic maps and hydraulic cross sections. New hydrological insights for the RegionThree flow domains were separated and characterized. Namely, i) the uppermost topography-driven flow systems, which penetrate only a few hundred meters, ii) a deep overpressured regime below 1600–2100 m depth, which drives fluids upward; and iii) a newly identified transition zone between the former two, which gains its energy from overpressure dissipation and contains non-renewable water resources. Topography-driven flow systems and discharge areas of the transition zone, where its upwelling saline water contributes to surface salinization, are parts of the CZ. Discharge areas of the transition zone cover about 50% of the Great Hungarian Plain. The overpressured system can only influence the CZ through the transition zone. The approach and methodology can be used in any terrestrial sedimentary basin where a deep overpressured regime exists.

Design and optimization of slot number in supercooled vapor suction in steam turbine blades for reducing the wetness

An enormous amount of the world's electricity is produced by steam power plants. The existence of a liquid phase in the steam turbine results in an efficiency drop and mechanical losses, which are very expensive due to its high price. Creating a suction slot in the stationary blade of the steam turbine for sucking the supercooled vapor and wetness fraction in the low-pressure stages of the turbine is one of the most effective methods of preserving the rotor blades against corrosion and erosion. The Eulerian-Eulerian method is employed for modeling the condensing flow. In the current research, the impact of the suction slot location and number on the surfaces of the turbine blades is investigated on condensation, total suction ratio, liquid suction ratio, average droplet radius, and condensation losses. Optimization is performed by the TOPSIS algorithm. Based on the results, the location and number of the suction slots affect the flow pattern in the turbine blade. Creating a suction slot in the optimal case has a total suction ratio of 4.1 %, which decreases the liquid suction ratio, condensation losses, and average droplet radius, respectively, by 5.28 %, 6 %, and 25 % compared to the original case.

Effect of Air Parameters on LiCl-H2O Film Flow Behavior in Liquid Desiccant Systems

The wettability and stability of a solution’s film on the filler surface are the key factors determining heat and mass transfer efficiency in liquid desiccant air conditioning systems. Therefore, this study investigates the effects of different air parameters on the flow behavior of a lithium chloride solution’s film. The effects of air velocity, air flow pattern, and pressure on the wettability and critical amount of spray are discussed. The results show that the main mechanism by which the air velocity affects the wettability is that the shear stress generated by the direction of the air velocity disperses the direction of the surface tension and weakens its effect on the liquid film distribution. In addition, in the counter flow pattern, the air flow blocks the liquid film from spreading longitudinally and destroys the stability of the liquid film at the liquid outlet, which increases the critical amount of spray. The pressure distribution is similar under different operating pressures when the flow is stable; thus, pressure has little effect on wettability. The simulation results under 8 atm are compared with the experimental results. It is found that the sudden increase in the amount of moisture removal when the amount of spray changes from 0.05 to 0.1 m3/(m·h) in the experiment is caused by the change in the liquid film flow state. In addition, the results show that within the range of air flow parameters for the liquid desiccant air conditioning system, air flow shear force is not the main factor affecting the stability of the solution’s film, and there is no secondary breakage of the solution’s film during the falling-film flow process.

Traffic flow prediction: A 3D adaptive multi‐module joint modeling approach integrating spatial‐temporal patterns to capture global features

AbstractThe challenges in citywide traffic flow are intricate, encompassing various factors like temporal and spatial dependencies, holidays, and weather. Despite the complexity, there are still research gaps in effectively incorporating these spatio‐temporal relations through deep learning. Addressing these gaps is crucial for tackling issues such as traffic congestion, public safety, and efficient traffic management within cities. This paper underscores notable research gaps, including the development of models capable of handling both local and global traffic flow patterns, integrating multi‐modal data sources, and effectively managing spatio‐temporal dependencies. In this paper, we proposed a novel model named 3D spatial–temporal‐based adaptive modeling graph convolutional network (3D(STAMGCN)) that addresses for traffic flow data in better periodicity modeling. In contrast to earlier studies, 3D(STAMGCN) approaches the task of traffic flow prediction as a periodic residual learning problem. This is achieved by capturing the input variation between historical time segments and the anticipated output for future time segments. Forecasting traffic flow, as opposed to a direct approach, is significantly simpler when focusing on learning more stationary deviations. This, in turn, aids in the training of the model. Nevertheless, the networks enable residual generation at each time interval through learned variations between future conditions and their corresponding weekly observations. Consequently, this significantly contributes to achieving more accurate forecasts for multiple steps ahead. We executed extensive experiments on two real‐world datasets and compared the performance of our model to state‐of‐the‐art (SOTA) techniques.

Numerical Calculation of Gas–Liquid Two-Phase Flow in Tesla Valve

In this paper, the gas–liquid two-phase flow within a Tesla valve under zero-gravity conditions is numerically studied. Based on the VOF model and the inlet two-phase separation method, the forward and reverse flow patterns and pressure drop changes in a Tesla valve at different inlet velocities were analyzed. At an inlet velocity of 0.1–0.2 m/s, the flow pattern was slug flow, the bubbles were evenly distributed in different positions in the Tesla valve, and the velocity difference between the main pipe and the arc branch pipe was small. When the inlet velocity was 0.4 m/s, the main flow pattern was annular flow, and there was a phenomenon of gas–liquid phase separation through different flow channels, which was related to centrifugal force. At an inlet velocity of 0.6–0.8 m/s, bubbly flow and slug flow coexisted, which was related to the uneven velocity. In the study range, the difference in the forward and reverse pressure drops of two-phase flow was smaller than that of single-phase flow, and the two-phase diodicity decreased first and then increased with the change in inlet velocity, reaching minimum values of 0.78 at 0.2 m/s and 1.44 at 0.8 m/s.

Two-phase flow visualization in direct ammonia fuel cells

In this work, the two-phase flow occurring in a direct ammonia fuel cell is visualized to understand the interactions between the complex physico-chemical processes involving in fuel cells. Experimental observations indicate that the appearance, growth, coalescence, detachment and sweeping processes of nitrogen bubbles occur periodically. Besides, the effect of operating conditions on the behavior of two-phase flow as well as cell performance is explored. Experimental results demonstrate that numerous nitrogen gas slugs are generated in the anode flow channel when the fuel flow rate is lower than 1.0 mL min−1, which seriously blocks the transport of NH3 and OH− to the anode catalyst layer. At an operating temperature of 95 °C, the flow channels are almost completely occupied by gases (a mixture of ammonia, nitrogen, and water), even under open-circuit conditions (a mixture of ammonia and water). In addition, the effects of using parallel and serpentine flow fields on the behaviour and flow patterns of two-phase flows are explored. Experimental results reveal that the parallel flow field is prone to clogging local flow channels by slugs. Moreover, the removal frequency of bubbles in the parallel design is significantly slower than that in the serpentine one.

Combining Computational Fluid Dynamics and Experimental Data to Understand Fish Schooling Behavior.

Understanding the flow physics behind fish schooling poses significant challenges due to the difficulties in directly measuring hydrodynamic performance and the three-dimensional, chaotic, and complex flow structures generated by collective moving organisms. Numerous previous simulations and experiments have utilized computational, mechanical, or robotic models to represent live fish. And existing studies of live fish schools have contributed significantly to dissecting the complexities of fish schooling. But the scarcity of combined approaches that include both computational and experimental studies, ideally of the same fish schools, has limited our ability to understand the physical factors that are involved in fish collective behavior. This underscores the necessity of developing new approaches to working directly with live fish schools. An integrated method that combines experiments on live fish schools with computational fluid dynamic (CFD) simulations represents an innovative method of studying the hydrodynamics of fish schooling. CFD techniques can deliver accurate performance measurements and high-fidelity flow characteristics for comprehensive analysis. Concurrently, experimental approaches can capture the precise locomotor kinematics of fish and offer additional flow information through particle image velocimetry (PIV) measurements, potentially enhancing the accuracy and efficiency of CFD studies via advanced data assimilation techniques. The flow patterns observed in PIV experiments with fish schools and the complex hydrodynamic interactions revealed by integrated analyses highlight the complexity of fish schooling, prompting a reevaluation of the classic Weihs model of school dynamics. The synergy between CFD models and experimental data grants us comprehensive insights into the flow dynamics of fish schools, facilitating the evaluation of their functional significance and enabling comparative studies of schooling behavior. In addition, we consider the challenges in developing integrated analytical methods and suggest promising directions for future research.

Identification of congenital aortic valve malformations in juvenile natriuretic peptide receptor 2 deficient (Npr2+/-) mice using high frequency ultrasound.

Mouse models of congenital aortic valve malformations are useful for studying disease pathobiology, but most models have incomplete penetrance (e.g., ~2 to 77% prevalence of bicuspid aortic valves (BAVs) across multiple models). For longitudinal studies of pathologies associated with BAVs and other congenital valve malformations, which manifest over months in mice, it is operationally inefficient, economically burdensome, and ethically challenging to enroll large numbers of mice in studies without first identifying those with valvular abnormalities. To address this need, we established and validated a novel in vivo high frequency (30 MHz) ultrasound imaging protocol capable of detecting aortic valvular malformations in juvenile mice. Fifty natriuretic peptide receptor 2 heterozygous mice on a low density lipoprotein receptor deficient background (Npr2+/-;Ldlr-/-; 32 male, 18 female) were imaged at 4- and 8-weeks of age. Fourteen percent of the Npr2+/-;Ldlr-/- mice exhibited features associated with aortic valve malformations, including: i) abnormal trans-aortic flow patterns on color Doppler (recirculation and regurgitation); ii) peak systolic flow velocities distal to the aortic valves reaching or surpassing ~1250 mm/s by pulsed wave Doppler; and iii) putative fusion of cusps along commissures and abnormal movement elucidated by 2D imaging with ultra-high temporal resolution. Valves with these features were confirmed by ex vivo gross anatomy and histological visualization to have thickened cusps, partial fusions, or Sievers type 0 bicuspid valves. This ultrasound imaging protocol will enable efficient, cost-effective and humane implementation of studies of congenital aortic valvular abnormalities and associated pathologies in a wide range of mouse models.

Investigation of the Effect of Heat Transfer during Friction Stir Welding (FSW) of AZ80A Mg Alloy Plates using a Pin Tool by Conducting Finite Elements Analysis

Friction stir welding (FSW) is an innovative solid-state welding process that has attracted substantial attention due to its potential for combining problematic materials such magnesium alloys, such as AZ80A. In order to better understand the impact of heat transport during FSW of AZ80A magnesium alloy plates using a pin tool, this study used finite element analysis (FEA). The welding process's thermal features, such as temperature distribution, thermal stresses, and material flow patterns, are the major focus of this analysis. The first step of the study is to conduct a comprehensive literature evaluation to lay a firm groundwork and pinpoint knowledge gaps. The thermal conductivity, specific heat, density, and mechanical characteristics of AZ80A magnesium alloy are measured and recorded as part of the material characterisation process. To ensure an exact simulation of real-world welding circumstances, a comprehensive 3D model of the welding setup is built, including the AZ80A magnesium alloy plates and the pin tool. In order to accurately record temperature variations, a tiny mesh is used, particularly in the welding zone. By include boundary conditions that mimic the real-world welding characteristics, such as the rotation of the pin tool and the clamping or fixturing of the plates, finite element analysis is used to model the FSW procedure. To simulate the heat input produced by FSW, a heat source or heat production model is used.

Quantification of steal flow on Doppler ultrasound in patients with arteriovenous graft for hemodialysis.

Arteriovenous fistula (AVF) is the preferred treatment for long-term hemodialysis patients to allow reliable vascular access. Arteriovenous graft (AVG) is monitored using Doppler sonography to check a vessel's condition and predict complications such as steal syndrome. In this study, we developed an analysis algorithm and method to quantify steal syndrome using Doppler sonography. Doppler sonography was used to determine the pattern of anterograde and retrograde flow. The ratio of blood volumes was calculated with a vision analysis software. First, performance of the developed algorithm was validated by comparing it with commercial Doppler sonography data. Doppler sonography was performed for an artificial vessel to analyze the steal flow. A total of 58 patients with steal flow were enrolled in this study. Of these patients, 23 did not have a difference in fingertip temperature between both sides. The median difference in temperature of 35 patients was 0.8°C (range, 0.3-1.9°C). The ratio of retrograde flow volume/antegrade flow volume in patients with the presence of temperature difference was significantly higher compared to that in patients without the temperature difference (p < .001). The ROC curve for the difference in flow volume had an AUC of 0.770. The optimal cutoff of difference in the flow volume between the two groups was 0.24 (sensitivity of 91.4 % and specificity of 52.2%). The flow volume difference was significantly positively correlated to temperature difference (r = 0.487, p < .003). Our algorithm could measure steal flow volume of a bidirectional waveform by antegrade arterial flow and retrograde reversal flow.

Study on the adsorption mechanism of droplets on parallel flow interfaces in U-shaped microchannels

The adsorption process of deionized water droplets on the parallel flow interface between ionic liquid [BMIM][PF6] and mineral oil is investigated in U-shaped microchannels. Three flow patterns of the efficient adsorption, moderate adsorption, and inefficient adsorption based on the adsorption efficiency were observed. The dimensionless droplet length D/W > 0.72 is conducive to efficient adsorption of droplets. When droplet shape factor β > 1, the lubrication pressure generated within the curved liquid film has a more significant hindrance on the adsorption process, and the reverse lubrication pressure at the outlet of the liquid film is the main reason for the rupture of the liquid film. An increase in the velocity of phase interface A-B movement can lead to an increase in lubrication pressure. A correlation model is established between the dimensionless adsorption length L* and the dimensionless adsorption time t*, and the influence of Laplace pressure is analyzed.

Phase Distribution of Gas–Liquid Slug–Annular Flow in Horizontal Parallel Micro-Channels

As a transitional flow pattern, slug–annular flow occurs over a wide range of operating conditions in micro-channels while its distribution in parallel micro-channels has not been well characterized. Herein, we conducted an experiment to study the phase distribution of slug–annular flow in parallel micro-channels. The test section consists of a header with a diameter of 0.48 mm and six branch channels with a diameter of 0.40 mm. Nitrogen and 0.03 wt% sodium dodecyl sulfate (SDS) solution were used as the test fluids. It was found that the phase distribution of the slug–annular flow was unstable and the duration of the varying process showed regularity with different inlet conditions. Increasing the liquid superficial velocity facilitated the liquid phase to flow into channels at the fore part of the header, while the channels at the rear part of the header were more supplied with liquid as the gas superficial velocity, volume fraction of gas, and volume flow rate increased. Furthermore, the results indicated that the channels located at the rear part of the header experienced a pronounced enhancement in the supply of both the liquid and gas phases, with the spacing between the branches increasing. A predictive correlation was formulated to ascertain the distribution of the liquid phase within slug–annular flow across parallel micro-channels.

How does the daily regulation hydropower station reduce the hydrological regime impact? A case study in upper Yellow River

ABSTRACT The regulation of hydropower stations has profoundly altered river flow patterns. While studies have extensively assessed the impact of large or multiyear regulated hydropower stations on hydrological regimes using indicators of hydrological alteration (IHA) and range of variability approach (RVA), the impact of daily regulation hydropower stations has received comparatively less attention. This study aims to evaluate the influence of daily regulation hydropower stations on hydrological regime changes. The analysis focuses on the upper Yellow River region in China, which houses cascade hydropower stations, utilizing daily runoff data from the Guide hydrological station spanning from 1954 to 2020. The Mann–Kendall test revealed 27 of 32 IHAs with a significant trend when considering the operation of the multiyear regulated hydropower station (Longyangxia). However, this number decreased to 18 IHAs when daily regulation hydropower stations (Laxiwa and Nina) were included. Evaluation using the RVA method indicated that only 46.87% of IHAs showed high alterations from the natural regime when considering the operation of daily regulation hydropower stations, a decrease from 75.00% when solely considering Longyangxia. Findings suggest that daily regulation hydropower stations can effectively mitigate the adverse effects of multiyear regulated hydropower stations, bringing the hydrological regime closer to a natural state.

Analysis of thermal videos for detection of lie during interrogation

The lie-detection tests are traditionally carried out by well-trained experts using polygraph machines. However, it is time-consuming, invasive, and, overall, a cumbersome process, not admissible by the court of law. Trained individuals can easily flaunt these tests. On the other hand, facial thermal imaging could be more effective as it is noninvasive and could be a stealth method of tracking the facial blood flow patterns, which have been proven to detect deceit. This paper presents a method based on facial thermal imaging to detect deception in human subjects. The major issue in such research is the lack of proper real-life databases which emulate the crime scenes. In this paper, first, we have developed a database based on almost real-life theft incidents with due diligence using isolated subjects over a period of time at a government hospital under the plea of free health checkup. The experiment has been conducted at Midnapore Medical College and Hospital, West Bengal, India, with proper ethical committee approval. The participants are selected at the behest of the police department with habitual crime records. Most of them have been repeatedly charged with petty crimes of pick-pocketing and stealing. They are invited individually at different instances of time under the plea of medical checkup where they have been enticed to steal cash. It is followed by a two-stage process, a friendly interaction followed by a slightly tougher interrogation. Their forehead and periorbital region skin surface temperature are recorded by a hidden thermal camera. Upon analysis, conspicuous changes in the temperature profile and blood flow pattern have been observed in the individuals who stole money and those who did not.

UNLOADING STRATEGIES IN ADVANCED HEART FAILURE: THE ROLE OF POINT–OF–CARE ULTRASOUND

Abstract Patients with advanced heart failure (HF) exhibit repeated episodes of fluid retention (pulmonary and/or systemic congestion) and/or of reduced cardiac output at rest (peripheral hypoperfusion) requiring hospitalization along with diuretics/inotropic infusion. Pulmonary artery catheter (PAC) provides haemodynamic information to guide the therapy, however PAC is invasive and time consuming. Point of care ultrasonography (POCUS) represents a promising diagnostic tool at the bedside, performed by the physician to assess non invasively volume status, cardiac output and monitor decongestive therapy. Recently, abnormal Doppler flow patterns due to increased right atrial pressure have been described in the hepatic, portal, and intrarenal veins. Thus, the recent venous excess ultrasound score (VExUS) has been used to describe fluid status along with the diameter and collapsibility of the inferior vena cava. Furthermore HF patients show high filling pressures, therefore they often require mechanical or medical unloading. In this regard, point of care ultrasound represents a valid tool to monitor non invasively the filling pressures and the responsiveness to those treatments. Here, we illustrate with several clinical cases where VExUS/POCUS were pivotal in clinical management.

Performance analysis of a gas turbine engine with intercooling and regeneration process - Part 1

Abstract Auxiliary systems, such as regeneration and intercooling, have been integrated with the primary gas generator to improve power production and fuel economy in modern gas turbine power plants. Implementing these techniques in turbine engines is challenging due to size, weight, and complex flow patterns. A solution is to use a turboprop engine with a smaller mass flow rate and simpler gas paths. The current study involves the numerical analysis of performance parameters namely, specific power (SP), thermal efficiency (η), and enthalpy based specific fuel consumption (EBSFC) of a turboprop engine using thermodynamic parameters namely, pressure ratio (PR), nozzle pressure ratio (NPR), turbine inlet temperature (TIT), regeneration efficiency (R), and intercooling efficiency (E). The results prove that the introduction of regeneration and intercooling showed significant improvement in the power developed, and reduced fuel consumption.

Effect of Salt Solution Tracer Dosage on the Transport and Mixing of Tracer in a Water Model of Asymmetrical Gas-Stirred Ladle with a Moderate Gas Flowrate

In previous research simulating steelmaking ladles using cold water models, the dosage/volume of the salt tracer solution is one of the factors that has been overlooked by researchers to a certain extent. Previous studies have demonstrated that salt tracers may influence the flow and measured mixing time of fluids in water models. Based on a water model scaled down from an industrial 130-ton ladle by a ratio of 1:3, this study investigates the impact of salt tracer dosage on the transport and mixing of tracers in the water model of gas-stirred ladle with a moderate gas flow rate. A preliminary uncertainty analysis of the experimental mixing time is performed, and the standard deviations were found to be less than 15%. It was observed in the experiments that the transport paths of tracers in the ladle can be classified into two trends. A common trend is that the injected salt solution tracer is asymmetrically transported towards the left sidewall of the ladle by the main circulation. In another trend, the injected salt solution tracer is transported both by the main circulation to the left side wall and by downward flow towards the gas column. The downward flow may be accelerated and become a major flow pattern when the tracer volume increases. For the dimensionless concentration curve, the sinusoidal type, which represents a rapid mixing, is observed at the top surface monitoring points, while the parabolic type is observed at the bottom monitoring points. An exception is the monitoring point at the right-side bottom (close to the asymmetric gas nozzle area), where both sinusoidal-type and parabolic-type curves are observed. Regarding the effect of tracer volume on the curve and mixing time, the curves at the top surface monitoring points are less influenced but curves at the bottom monitoring points are noticeably influenced by the tracer volume. A trend of decreasing and then increasing as the tracer volume increases was found at the top surface monitoring points, while the mixing times at the bottom monitoring points decrease with the increase in the tracer volume.

Understanding two‐phase flow behaviour: CFD Assessment of silicone oil–air and water–air in an intermediate vertical pipe

AbstractThe study utilized computational fluid dynamics (CFD) simulations employing the volume of fluid (VOF) model to analyze silicone oil–air flows in a vertical pipe with a diameter of 67 mm. Various structured meshes ensured grid independence, and the k‐ε realizable model addressed turbulence. The analysis characterized bubbly to annular flows, involving the evaluation of flow patterns, radial void fractions, void fraction time series, probability density functions (PDFs), power spectral densities (PSDs), and mean void fractions. Results indicated a transition from bubbly to annular flow with increasing gas velocities and notable changes in radial void fraction profiles. Void fraction exhibited significant variations with distinct flow patterns at constant liquid velocity. PDFs identified flow regimes, and PSDs revealed frequency patterns. CFD results were validated against experiments, demonstrating good agreement. The validated CFD model was utilized to investigate radial gas velocities and pressure drops, revealing a shift from uniform velocity distributions to irregular patterns and a decrease in total pressure drop with an increase in gas superficial velocity. The model was also applied to a water‐air system to explore two‐phase flow behaviour. The impact of superficial gas velocity on flow patterns and radial void fractions was studied through numerical analysis and compared with the silicone oil‐air system. Results showed that at extreme gas velocities (0.06 and 5.53 m/s), silicone oil exhibited bubbly and annular flows, while water displayed cap bubbly and churn flows. The significant variation in radial void fraction at these velocities emphasized the impact of fluid properties.

Topological defects in multi-layered swarming bacteria.

Topological defects, which are singular points in a director field, play a major role in shaping active systems. Here, we experimentally study topological defects and the flow patterns around them, that are formed during the highly rapid dynamics of swarming bacteria. The results are compared to the predictions of two-dimensional active nematics. We show that, even though some of the assumptions underlying the theory do not hold, the swarm dynamics is in agreement with two-dimensional nematic theory. In particular, we look into the multi-layered structure of the swarm, which is an important feature of real, natural colonies, and find a strong coupling between layers. Our results suggest that the defect-charge density is hyperuniform, i.e., that long range density-fluctuations are suppressed.

Air Traffic Flow Prediction with Spatiotemporal Knowledge Distillation Network

Accurate air traffic flow prediction assists controllers formulate control strategies in advance and alleviate air traffic congestion, which is important to flight safety. While existing works have made significant efforts in exploring the high dynamics and heterogeneous interactions of historical air traffic flow, two key challenges still remain. (1) The transfer patterns of air traffic are intricate, subject to numerous constraints and limitations such as controllers, flight regulations, and other regulatory factors. Relying solely on mining historical traffic evolution patterns makes it difficult to accurately predict the constrained air traffic flow. (2) Weather conditions exert a substantial influence on air traffic, making it exceptionally difficult to simulate the impact of external factors (such as thunderstorms) on the evolution of air traffic flow patterns. To address these two challenges, we propose a Spatiotemporal Knowledge Distillation Network (ST-KDN) for air traffic flow prediction. Firstly, recognizing the inherent future insights embedded within flight plans, we develop a “teacher-student” distillation model. This model leverages the prior knowledge of upstream-downstream migration patterns and future air traffic trends inherent in flight plans. Subsequently, to model the influence of external factors and predict air traffic flow disturbed by thunderstorm weather, we propose a student network based on the “parallel-fusion” structure. Finally, employing a feature-based knowledge distillation approach to integrate prior knowledge from flight plans and extract meteorological features, our method can accurately capture complex and constrained spatiotemporal dependencies in air traffic and explicitly model the impact of weather on air traffic flow. Experimental results on real-world flight data demonstrate that our method can achieve better prediction performance than other state-of-the-art comparison methods, and the advantages of the proposed method are particularly prominent in modeling the complicated transfer pattern of air traffic and inferring nonrecurrent flow patterns.