Current status and prospects of research on marine atmospheric duct characteristics and their applications

【Significance】 Maritime radar, wireless communication and electronic countermeasure systems are foundational to ocean situational awareness, offshore operations and maritime security. Their performance is deeply shaped by atmospheric ducts—super-refractive stratifications caused by anomalous vertical profiles of temperature, humidity and pressure in the marine atmospheric boundary layer. Ducts trap electromagnetic waves to enable low-loss over-the-horizon propagation, with a dual-sided impact: they can extend detection and communication ranges far beyond the line of sight, but also induce radar blind zones, ranging errors, and abnormal sea clutter. With the rapid development of over-the-horizon radar, early warning systems and 6G maritime communication architectures, atmospheric duct research has grown increasingly critical. This paper systematically reviews global advances in marine atmospheric duct research, covering theoretical modeling, spatiotemporal distributions, monitoring and inversion techniques, numerical forecasting and engineering applications. It also presents the research progress of the Northwestern Polytechnical University team and outlines key future directions.【Progress】 Marine atmospheric ducts fall into three categories with distinct formation mechanisms: evaporation ducts, surface-based ducts and elevated ducts. Driven by sea surface evaporation, evaporation ducts exist almost ubiquitously over oceans, typically below 40 m high with a global average of 13 m, and exert the strongest influence on near-surface electromagnetic systems. Surface-based and elevated ducts, induced by warm air advection, subtropical high subsidence and frontal processes, occur at higher altitudes with greater spatiotemporal variability. In theoretical modeling, evaporation duct diagnostic models have evolved from empirical formulations to physics-based bulk frameworks. Early models such as the Paulus-Jeske model offer high computational efficiency but may show reduced accuracy under stable stratification and coastal conditions. Physics-based LKB-derived models (NRL, NWA, BYC, NAVSLaM) improve fidelity via air-sea flux coupling. Field validations in the South China Sea reveal strong stability-dependent performance: the NPS model excels under unstable conditions, while the P-J model suits stable stratification better. Data-driven methods including LSTM, TCN-GRU and deep forest serve as complements, achieving small prediction errors by integrating multi-dimensional meteorological inputs. For electromagnetic wave propagation, the parabolic equation method with split-step Fourier transformation remains the mainstream for handling range-dependent refractivity and complex terrain. Ray tracing, waveguide mode theory and hybrid models like the Advanced Propagation Model address complementary scenarios, while deep learning models have recently improved path-loss prediction accuracy with reduced computational cost. Software suites such as PETOOL and the U.S. Navy′s AREPS support engineering applications. Spatiotemporal distribution studies using reanalysis datasets and in-situ measurements have characterized duct features across major sea areas. Duct height exhibits pronounced seasonal and diurnal cycles modulated by air-sea temperature difference, wind speed and relative humidity. Extreme weather events leave distinct signatures: typhoons feature an “evaporation duct eye” with suppressed height in the eye and enhanced ducting in the eyewall; cold surges significantly elevate duct height via strong wind mixing; sea fog, oceanic fronts and rainfall also induce characteristic structural modulations. Propagation studies confirm that path loss is jointly governed by duct height, horizontal inhomogeneity, maritime obstacles, operating frequency and antenna height, with optimal frequency and antenna height varying dynamically with duct conditions. Four duct environment acquisition approaches have been developed with complementary strengths. Direct measurement delivers high-fidelity profiles but entails high cost and limited spatial coverage. Model estimation enables large-area continuous diagnosis but is constrained by boundary layer parameterization biases. Inversion techniques provide environment information directly aligned with electromagnetic system performance, with recent advances in multi-frequency joint inversion and phase-coherent retrieval substantially improving accuracy. For forecasting, mesoscale numerical weather prediction models coupled with duct diagnostic models form the primary framework, while emerging AI weather models show promise for short-term nowcasting when integrated with physical duct models. Comprehensive field experiments are critical for model validation and mechanistic understanding. The RED experiment, TAPS and the landmark CASPER program established benchmark datasets. The recent REDSAW campaign specifically targets stable boundary layer regimes, filling long-standing gaps in low-turbulence duct characterization. In engineering applications, ducts can extend radar ranges to over 100 km but also cause blind zones and positioning errors. For maritime communications, evaporation ducts support multi-Mbps data rates over tens to hundreds of kilometers, offering a complementary backhaul solution for offshore IoT systems. The Northwestern Polytechnical University team has conducted systematic long-term research on marine atmospheric ducts in China. It has revealed the quantitative correlation between evaporation ducts and oceanic evaporation, developed empirical and data-driven estimation models, proposed multiple inversion approaches, established climatological distributions and propagation statistics for China′s marginal seas, and developed forecasting systems, software tools and over-the-horizon communication systems validated under favorable duct conditions.【Conclusions and Prospects】 Marine atmospheric duct research has built a solid foundation in theory, observation and applications, but still faces challenges in long-term open-ocean observation, high-accuracy forecasting under complex conditions, and system-level engineering integration. Future work should prioritize five directions: refining theoretical models with multi-scale air-sea interaction processes; constructing an integrated air-space-sea observation network; improving forecasting resolution via coupled models and physics-informed artificial intelligence; developing environment-adaptive radar and communication systems; and advancing standardization to facilitate cross-institutional data sharing. Overall, the field is transitioning from empirical characterization to intelligent, system-level integration, which will underpin next-generation maritime electromagnetic systems.