This novel approach, termed TCI–KAN, represents a fusion of deep learning with interpretable neural architectures, specifically leveraging Kolmogorov–Arnold networks (KANs) alongside a dynamic predictor pruning optimization module. The architecture of TCI–KAN breaks away from conventional deep learning systems, enhancing both efficiency and interpretability in capturing complex atmospheric dynamics influencing cyclonic intensification. The framework is structured around three primary modules: a predictor pruning optimization module that intelligently selects the most influential input parameters, a neural network optimization module fine-tuning the model’s learning capability, and a prediction module that generates precise intensity forecasts.

The driving innovation behind TCI–KAN lies in its ability to prune a vast pool of potential predictors down to a concise subset that significantly impacts the prediction of tropical cyclone intensity. From an initial collection of 317 predictors—variables ranging from oceanic thermodynamics to atmospheric conditions—this pruning mechanism distills the inputs to just 15 high-impact features. This reduction not only streamlines computational complexity but also enhances model interpretability, a crucial advantage over typical black-box deep learning methods that often struggle to elucidate their decision-making processes.

Testing the TCI–KAN framework on historical cyclone data revealed breakthrough performance, particularly in six-hour intensity forecasts where it achieved a mean absolute error (MAE) of only 2.85 knots. This result marks a significant leap forward, outperforming current operational forecasts by 31 percent and exceeding the accuracy of both single and hybrid deep learning models by 13 and 6 percent, respectively. Such precision improvements are instrumental in providing coastal regions and emergency planners with reliable, timely warnings that can save lives and reduce economic losses.

Beyond accuracy, TCI–KAN demonstrates remarkable versatility and robustness across different ocean basins and tropical cyclone categories. While exhibiting the highest fidelity in the eastern Pacific—a region characterized by a particular set of environmental influences—the model maintains strong predictive capabilities in other areas, adjusting to varying storm intensities. Notably, the framework’s uncertainty in prediction increases moderately with escalating cyclone intensity, reflecting inherent challenges in modeling extreme atmospheric phenomena but still offering superior confidence compared to existing methods.

At the heart of TCI–KAN’s success is the interpretability offered by Kolmogorov–Arnold networks, which differ from traditional deep neural networks by decomposing complex nonlinear mappings into simpler functions. This mathematical foundation allows researchers to better understand and trust the internal workings of the model—a significant stride in applying artificial intelligence in operational meteorology where transparency is essential. The dynamic predictor pruning module enhances this by continuously optimizing the feature set, ensuring that the model adapts to evolving atmospheric conditions and data availability.

Professor Wei Zhong underscores the broader implications of this research, emphasizing that the integration of data-driven techniques with physical mechanisms heralds a new era in meteorological forecasting. “TCI–KAN not only pushes the boundary of forecasting accuracy but also bridges the gap between interpretable machine learning and the traditionally physical mechanism-based methods,” he stated. This fusion can pave the way toward next-generation forecasting systems that balance empirical data insights with robust atmospheric science principles.

The practical implications for disaster management agencies and meteorological services worldwide are profound. Enhanced six-hour intensity forecasts can enable better allocation of resources, refined evacuation planning, and more targeted warnings that reduce unnecessary economic disruptions. Furthermore, the model’s adaptability across regions suggests it could be globally adopted and tailored to local cyclone characteristics, representing a universal tool in the fight against tropical cyclone hazards.

This research also contributes to the ongoing discourse about the role of artificial intelligence in environmental and geophysical sciences. By demonstrating that deep learning models can be both highly accurate and interpretable, TCI–KAN challenges the assumption that sophisticated AI methods must remain opaque. Instead, it illustrates a path forward where explainability complements performance—an essential balance for operational deployment and scientific advancement alike.

The foundation of this work rests heavily on rigorous mathematical optimization, feature selection techniques, and neural network training algorithms that are intricately designed to capture the dynamic and chaotic nature of tropical cyclones. The pruning optimization reduces input redundancy and noise, focusing computational power and model attention on the most relevant physical indicators, such as sea surface temperatures, wind shear parameters, and moisture content profiles—elements known to critically influence storm evolution.

Developed through meticulous experimentation and validation against historical basin-wide datasets, TCI–KAN’s deployment is timely given the increasing threat of intense storms fueled by climate change. As ocean temperatures rise and more variable atmospheric conditions emerge, predictive tools must evolve in tandem to safeguard vulnerable populations and infrastructure more effectively.

Keyun Li, a master’s student and the first author of the publication, played a pivotal role in designing and testing the TCI–KAN framework under Professor Zhong’s guidance. Their collaborative efforts were supported by the National Natural Science Foundation of China, reflecting a national commitment to advancing meteorological sciences through cutting-edge interdisciplinary research spanning physics, computer science, and atmospheric dynamics.

Published in the reputable journal Atmospheric and Oceanic Science Letters, this study sets a new benchmark for tropical cyclone intensity prediction research. It invites further exploration into the fusion of interpretable AI models with traditional forecasting methods, and is expected to influence future developments in the field, including real-time operational use and expanded applications to other extreme weather phenomena.

As the climate evolves and risks from tropical cyclones intensify, innovations like TCI–KAN represent a beacon of progress. They illustrate how the convergence of data science and atmospheric physics can lead to safer, smarter, and more responsive forecasting systems essential for the resilience of societies worldwide.

Subject of Research: Tropical cyclone intensity prediction using interpretable deep learning networks.

Article Title: Tropical cyclone intensity prediction based on Kolmogorov–Arnold networks with predictor pruning optimization

News Publication Date: 13-Aug-2025

Web References:
https://doi.org/10.1016/j.aosl.2025.100694

Image Credits: Keyun Li

Keywords: Tropical cyclones, Deep learning, Meteorology, Cyclone intensity prediction, Kolmogorov–Arnold networks, Predictor pruning optimization, Interpretability, Atmospheric science

Tropical cyclones originating in the Bay of Bengal pose a unique challenge to oceanographers and meteorologists alike, largely due to the Bay’s intricate monsoonal influences which create divergent oceanic conditions across seasons. Pre-monsoon cyclones, forming between March and May, and post-monsoon cyclones, developing from October to December, manifest starkly different thermal and salinity profiles that influence their physical structure and progression. Understanding these differences is pivotal in decoding the mechanisms behind cyclone-ocean feedbacks.

Central to the study is the observation that pre-monsoon cyclones evolve over warmer sea surface temperatures (SST), fluctuating between 29°C and 31°C. These elevated temperatures enhance the upper ocean’s thermal stratification and drive significant freshwater inflow from coastal rivers and tributaries during this season, leading to coastal freshening—an observable reduction in sea surface salinity. This freshening plays a crucial role in modulating mixed layer depth and subsequently affects the transfer of heat and momentum from the atmosphere to the ocean.

In stark contrast, the post-monsoon cyclones are found to develop over relatively cooler oceans, with the SST stabilizing between 28°C and 29°C. During this phase, the dominant force appears to shift from ocean surface temperature to atmospheric dynamics such as intensified rainfall and convection. The study highlights that post-monsoon cyclones are associated with heavier precipitation events which increase chlorophyll-a concentrations within the mixed layer. Elevated chlorophyll-a levels between 9.9 and 14.4 mg/m³ point to enhanced biological productivity triggered by nutrient upwelling, which is in turn stimulated by cyclone-induced Ekman transport mechanisms.

A defining feature of both cyclone types is their influence on the mixed layer depth (MLD). Pre-monsoon cyclones, with their higher SST and freshwater influx, tend to create shallower mixed layers near coastal zones. This stratified environment impedes vertical mixing and contributes to heat accumulation in the upper ocean, potentially amplifying storm intensity. Post-monsoon storms, conversely, generate deeper mixed layers due to increased wind-driven turbulence and nutrient entrainment, which have significant repercussions on marine ecosystems and biogeochemical cycles.

The authors pay particular attention to the role of sea surface salinity (SSS) variability. Pre-monsoon cyclones are linked to a stronger freshwater influence, as the runoff from monsoonal rivers dilutes the coastal salinity and influences buoyancy fluxes. This freshwater layer acts to stabilize the upper ocean, modifying the heat content available to cyclones. Post-monsoon cyclones, subjected to intense rainfall rather than freshwater inflow, experience surface freshening patterns with distinct spatial and temporal signatures, demanding nuanced observation approaches.

Another intriguing aspect illuminated by the study is the sea level pressure (SLP) distribution associated with these cyclones. Post-monsoon cyclones demonstrate lower central pressures and stronger wind fields, perhaps a consequence of sustained atmospheric convection and ocean-atmosphere coupling specific to the cooler SST regime. This interplay between ocean surface conditions and atmospheric forcing not only governs cyclone lifespan but also their potential for intensification or dissipation.

The research team underscores the temporal lag in oceanic responses to cyclone events. While some variables, like SST and SLP, display immediate changes, others such as biological productivity and nutrient distribution respond over longer time scales. This delay highlights the necessity for comprehensive, high-resolution ocean monitoring both during and after cyclone passage to fully capture the cascade of physical and biogeochemical processes at play.

Despite the robustness of these findings, the authors acknowledge limitations inherent to the study. The analysis is constrained by a relatively small sample size and focuses on a limited temporal window, which may not account for interannual variability or anomalies. Nevertheless, the results lay a solid foundation for future expansive research endeavors integrating real-time satellite observations and advanced ocean-atmosphere coupled models.

Foreseeing the implications of their work, the researchers advocate for the development of predictive tools that can utilize these seasonal oceanic signatures to enhance cyclone forecasting accuracy. Early warning systems, informed by differential ocean responses, could improve disaster preparedness not only in the Bay of Bengal but also in similar tropical cyclone-prone regions with marked seasonal monsoonal cycles.

This study exemplifies the critical need to integrate physical oceanography and atmospheric science to unravel the complexities of cyclone behavior in monsoon-dominated environments. The intricate feedbacks between ocean surface conditions and atmospheric processes are essential parameters that determine cyclone genesis, evolution, and potential destructiveness.

Contributors to this impactful research hail from the Bangladesh Oceanographic Research Institute’s Physical, Space, and Biological Oceanography Divisions as well as the Department of Oceanography at the University of Chittagong. Their collaborative efforts highlight the importance of regional expertise in addressing climate challenges relevant to vulnerable coastal communities.

In conclusion, the nuanced differentiation between pre- and post-monsoon cyclone interactions with the upper ocean unveiled by this study advances our understanding of cyclone dynamics in the Bay of Bengal. It also informs broader climate and oceanographic models that underpin global efforts to predict and mitigate the impacts of tropical cyclones in a changing climate, emphasizing the critical role that seasonality plays in modulating oceanic and atmospheric processes.

—
Subject of Research: Not applicable
Article Title: Upper Ocean Response Mechanisms to Pre-Monsoon and Post-Monsoon Cyclones in the Bay of Bengal
News Publication Date: 25-Aug-2025
Web References: http://dx.doi.org/10.34133/olar.0105
References: Chowdhury, S. U. M. B., Karmakar, A., Hoque, M. E., Hoque, M. M., Tahsin, T. H., & Chowdhury, S. (2025). Upper Ocean Response Mechanisms to Pre-Monsoon and Post-Monsoon Cyclones in the Bay of Bengal. Ocean-Land-Atmosphere Research.
Image Credits: Chowdhury, S. U. M. B., Karmakar, A., Hoque, M. E., Hoque, M. M., Tahsin, T. H., & Chowdhury, S. (2025).

Keywords

Air sea interactions, Ocean circulation, Meteorology