KAN-FIF: Spline-Parameterized Lightweight Physics-based Tropical Cyclone Estimation on Meteorological Satellite

2026-02-12

AI Summary

This study addresses the high computational cost and parameter redundancy of existing tropical cyclone intensity estimation models, which hinder deployment on edge devices and limit their ability to capture high-order nonlinear relationships among meteorological features. To overcome these challenges, we propose KAN-FIF, a lightweight multimodal framework that, for the first time, integrates Kolmogorov–Arnold Networks (KANs) with spline-based parameterization into this domain, synergistically combining CNNs, MLPs, and physics-guided mechanisms. Compared to the Phy-CoCo baseline, our model reduces parameters by 94.8% (0.99 MB vs. 19 MB), accelerates per-sample inference by 68.7% (2.3 ms vs. 7.35 ms), and achieves a 32.5% lower mean absolute error (MAE). Furthermore, it enables efficient edge deployment on the FY-4 satellite processor at 14.41 ms per sample.

Technology Category

• Machine Learning: Learning on the Edge & Model Compression
• Computer Vision: Multi-modal Vision
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG

Abstract

Tropical cyclones (TC) are among the most destructive natural disasters, causing catastrophic damage to coastal regions through extreme winds, heavy rainfall, and storm surges. Timely monitoring of tropical cyclones is crucial for reducing loss of life and property, yet it is hindered by the computational inefficiency and high parameter counts of existing methods on resource-constrained edge devices. Current physics-guided models suffer from linear feature interactions that fail to capture high-order polynomial relationships between TC attributes, leading to inflated model sizes and hardware incompatibility. To overcome these challenges, this study introduces the Kolmogorov-Arnold Network-based Feature Interaction Framework (KAN-FIF), a lightweight multimodal architecture that integrates MLP and CNN layers with spline-parameterized KAN layers. For Maximum Sustained Wind (MSW) prediction, experiments demonstrate that the KAN-FIF framework achieves a $94.8\%$ reduction in parameters (0.99MB vs 19MB) and $68.7\%$ faster inference per sample (2.3ms vs 7.35ms) compared to baseline model Phy-CoCo, while maintaining superior accuracy with $32.5\%$ lower MAE. The offline deployment experiment of the FY-4 series meteorological satellite processor on the Qingyun-1000 development board achieved a 14.41ms per-sample inference latency with the KAN-FIF framework, demonstrating promising feasibility for operational TC monitoring and extending deployability to edge-device AI applications. The code is released at https://github.com/Jinglin-Zhang/KAN-FIF.

Problem

• Tropical Cyclone Estimation
• Edge Deployment
• Model Efficiency
• Physics-based Modeling
• High-order Feature Interactions

Innovation

• Kolmogorov-Arnold Network
• spline-parameterized
• lightweight physics-based modeling
• tropical cyclone estimation
• edge AI deployment