Tag
This paper benchmarks six generative deep learning models for nowcasting aerosol particle size distributions from PM2.5 and meteorological data, finding that physics-guided flow matching performs best, with a reduced general dynamic equation constraint lowering CRPS by 5.5–11.5% at the training site.
Jeff Geerling reviews and tears down Ubiquiti's $99 UP-AirQuality vape detector and air quality sensor, finding its Sensirion SEN66 sensor delivers CO2 readings surprisingly close to higher-end monitors like the Aranet4 and AirGradient.
GAUDI proposes a conditional diffusion model that leverages missingness geometry for imputing contiguous missing blocks in air-quality time-series data, showing improved performance over existing methods.
The author purchased two air purifiers for their new home in Cyprus to improve air quality and reduce health risks from PM2.5 particles.
The University of Manchester has used NVIDIA's Earth-2 AI models to develop a forecasting system for air pollution across the UK, aiming to improve public health responses and policy decisions.
This paper introduces QHAdamW, a modified optimizer for artificial neural networks, applied to air quality forecasting in the Philippines, showing improved convergence and performance in predicting PM2.5 and PM10 levels.
Thick smoke from Canadian wildfires blankets cities from Chicago to Washington, DC, turning skies orange and triggering air quality alerts; the article highlights the role of climate change in worsening wildfire seasons and associated health risks.
The article discusses the severe health risks of wildfire smoke and provides guidance on using N95 masks, air purifiers, and monitoring air quality via Airnow.gov to stay safe.
Smoke from Canadian wildfires blankets the eastern US, causing hazardous air quality and prompting health warnings as climate change intensifies wildfire seasons.
OmniPM-Net is a neural process fusion model that combines discrete station forecasts from graph neural networks with gridded forecasts from chemical transport models to produce consistent PM10 predictions at both stations and grid cells, improving accuracy especially during dust storms.
This paper systematically evaluates time series foundation models (TSFMs) for forecasting extreme PM2.5 concentrations from wildfire smoke using a 12-year dataset from California. Results show that fully-trained recurrent baselines like BiLSTM outperform TSFMs, challenging the assumption that larger pretrained models dominate in environmental forecasting.
High CO2 levels in meeting rooms can significantly impair decision-making and cognitive performance, as shown in research. The author recommends using CO2 monitors and opening windows to improve meeting effectiveness.
This paper introduces low-cost High-Order Singular Value Decomposition (lcHOSVD), a tensor-based method for reconstructing high-dimensional environmental fields from sparse sensor measurements. Applied to urban flow and air-quality datasets, it achieves lower reconstruction errors and greater robustness to uneven sensor distributions compared to matrix-based approaches.
This paper proposes a confusion matrix-based graph construction method and a hybrid loss function for Graph Neural Networks to improve multi-site pollution prediction accuracy and interpretability, evaluated on real-world air pollution data.
This paper proposes MVG-KAN, a multi-view model integrating periodic-residual decomposition, a Geo-Wind Graph for wind-aware spatial dependencies, and a temporal KAN head for PM2.5 forecasting, achieving MAE 14.09 on Beijing data.
The Birdie Pro is an updated air quality monitor that uses a mechanical bird to indicate poor air quality, now with additional sensors and smart home integrations. It's available via Kickstarter.
Proposes M²FedAQI, a lightweight multimodal federated learning framework for air quality prediction across heterogeneous edge devices, achieving significant improvements over baselines on benchmark datasets.