TinyNina: A Resource-Efficient Edge-AI Framework for Sustainable Air Quality Monitoring via Intra-Image Satellite Super-Resolution

2026-04-06

AI Summary

This study addresses the challenge of limited spatial resolution in satellite remote sensing data—such as Sentinel-2—which hinders fine-grained monitoring of NO₂ pollution. The authors propose a lightweight, self-supervised super-resolution framework tailored for edge devices, leveraging only internal information from a single multispectral image without requiring external high-resolution labels. The method integrates wavelength-specific attention gating with depthwise separable convolutions, preserving spectral features sensitive to pollutants while using merely 51K parameters. Evaluated on 3,276 matched satellite–ground station samples, the approach achieves a mean absolute error (MAE) of 7.4 μg/m³, reduces computational overhead by 95%, and accelerates inference by 47× compared to EDSR and RCAN, thereby significantly advancing efficient and sustainable remote sensing for air quality monitoring.

Abstract

Nitrogen dioxide (NO$_2$) is a primary atmospheric pollutant and a significant contributor to respiratory morbidity and urban climate-related challenges. While satellite platforms like Sentinel-2 provide global coverage, their native spatial resolution often limits the precision required, fine-grained NO$_2$ assessment. To address this, we propose TinyNina, a resource-efficient Edge-AI framework specifically engineered for sustainable environmental monitoring. TinyNina implements a novel intra-image learning paradigm that leverages the multi-spectral hierarchy of Sentinel-2 as internal training labels, effectively eliminating the dependency on costly and often unavailable external high-resolution reference datasets. The framework incorporates wavelength-specific attention gates and depthwise separable convolutions to preserve pollutant-sensitive spectral features while maintaining an ultra-lightweight footprint of only 51K parameters. Experimental results, validated against 3,276 matched satellite-ground station pairs, demonstrate that TinyNina achieves a state-of-the-art Mean Absolute Error (MAE) of 7.4 $μ$g/m$^3$. This performance represents a 95% reduction in computational overhead and 47$\times$ faster inference compared to high-capacity models such as EDSR and RCAN. By prioritizing task-specific utility and architectural efficiency, TinyNina provides a scalable, low-latency solution for real-time air quality monitoring in smart city infrastructures.

Problem

• NO2 monitoring
• satellite super-resolution
• spatial resolution
• air quality assessment
• remote sensing

Innovation

• intra-image learning
• edge AI
• satellite super-resolution
• lightweight neural network
• NO2 monitoring