Ever growing populations in cities are associated with a major increase in road vehicles and air pollution. The overall high levels of urban air pollution have been shown to be of a significant risk to city dwellers. However, the impacts of very high but temporally and spatially restricted pollution, and thus exposure, are still poorly understood. Conventional approaches to air quality monitoring are based on networks of static and sparse measurement stations. However, these are prohibitively expensive to capture tempo-spatial heterogeneity and identify pollution hotspots, which is required for the development of robust real-time strategies for exposure control. Current progress in developing low-cost micro-scale sensing technology is radically changing the conventional approach to allow real-time information in a capillary form. But the question remains whether there is value in the less accurate data they generate. This article illustrates the drivers behind current rises in the use of low-cost sensors for air pollution management in cities, whilst addressing the major challenges for their effective implementation.

The rise of low-cost sensing for managing air pollution in cities

Mapping spatial distribution of particulate matter using Kriging and Inverse Distance Weighting at supersites of megacity Delhi

This article devises a photograph-based monitoring model to estimate the real-time PM2.5 concentrations, overcoming currently popular electrochemical sensor-based PM2.5 monitoring methods’ shortcomings such as low-density spatial distribution and time delay. Combining the proposed monitoring model, the photographs taken by various camera devices (e.g., surveillance camera, automobile data recorder, and mobile phone) can widely monitor PM2.5 concentration in megacities. This is beneficial to offering helpful decision-making information for atmospheric forecast and control, thus reducing the epidemic of COVID-19. To specify, the proposed model fuses Information Abundance measurement and Wide and Deep learning, dubbed as IAWD, for PM2.5 monitoring. First, our model extracts two categories of features in a newly proposed DS transform space to measure the information abundance (IA) of a given photograph since the growth of PM2.5 concentration decreases its IA. Second, to simultaneously possess the advantages of memorization and generalization, a new wide and deep neural network is devised to learn a nonlinear mapping between the above-mentioned extracted features and the groundtruth PM2.5 concentration. Experiments on two recently established datasets totally including more than 100 000 photographs demonstrate the effectiveness of our extracted features and the superiority of our proposed IAWD model as compared to state-of-the-art relevant computing techniques.

PM₂.₅ Monitoring: Use Information Abundance Measurement and Wide and Deep Learning

This paper presents an engine able to forecast jointly the concentrations of the main pollutants harming people's health: nitrogen dioxyde (NO2), ozone (O3) and particulate matter (PM2.5 and PM10, which are respectively the particles whose diameters are below 2.5 um and 10 um respectively). 
The forecasts are performed on a regular grid (the results presented in the paper are produced with a 0.5° resolution grid over Europe and the United States) with a neural network whose architecture includes convolutional LSTM blocks. The engine is fed with the most recent air quality monitoring stations measures available, weather forecasts as well as air quality physical and chemical model (AQPCM) outputs. The engine can be used to produce air quality forecasts with long time horizons, and the experiments presented in this paper show that the 4 days forecasts beat very significantly simple benchmarks. 
A valuable advantage of the engine is that it does not need much computing power: the forecasts can be built in a few minutes on a standard GPU. Thus, they can be updated very frequently, as soon as new air quality measures are available (generally every hour), which is not the case of AQPCMs traditionally used for air quality forecasting. 
The engine described in this paper relies on the same principles as a prediction engine deployed and used by Plume Labs in several products aiming at providing air quality data to individuals and businesses.

PlumeNet: Large-Scale Air Quality Forecasting Using A Convolutional LSTM Network

A Big Data and Artificial Intelligence Framework for Smart and Personalized Air Pollution Monitoring and Health Management in Hong Kong

The surveillance of air pollution is becoming a highly concerned issue for city residents and urban administrators. Fixed air quality stations as well as mobile gas sensors have been deployed for air quality monitoring but with sparse observations over the entire temporal-spatial space. Therefore, an inference algorithm is essential for comprehensive fine-grained air pollution sensing. Conventional physically-based models can hardly be applied to all the scenarios, while pure data-driven methods suffer from sampling bias and overfitting problems. This paper presents a hybrid algorithm for air pollution inference by guiding the data learning process with physical model. The quantitative combination of knowledge from observed dataset and a discretized convective-diffusion model is performed within a multi-task learning scheme. Evaluations show that, benefited from physical guidance, our hybrid method obtains higher extrapolation ability and more robustness, achieving the same performance with 1/8 sample amount and obtaining 31.9% less error in noisy synthesized environment. In a real-world deployment in Tianjin, our algorithm outperforms the pure data-driven model with 9.69% less inference error over a 9-day PM 2.5 data collection.

Guiding the Data Learning Process with Physical Model in Air Pollution Inference

Air pollution is a global health threat. Except static official air quality stations, mobile sensing systems are deployed for urban air pollution monitoring to achieve larger sensing coverage and greater sampling granularity. However, the data sparsity and irregularity also bring great challenges for pollution map recovery. To address these problems, we propose a deep autoencoder framework based inference algorithm. Under the framework, a partially observed pollution map formed by the irregular samples are input into the model, then an encoder and a decoder work together to recover the entire pollution map. Inside the decoder, we adopt a convolutional long short-term memory (ConvLSTM) model by revealing its physical interpretation with an atmospheric dispersion model, and further present a weather-related ConvLSTM to enable quasi real-time applications. To evaluate our algorithm, a half-year data collection was deployed with a real-world system on a coastal area including the Sino-Singapore Tianjin Eco-city in north China. With the resolution of 500 m x 500 m x 1 h, our offline method is proved to have high robustness against low sampling coverage and accidental sensor errors, obtaining 14.9% performance improvement over existing methods. Our quasi real-time model better captures the spatiotemporal dependencies in the pollution map with unevenly distributed samples than other real-time approaches, obtaining 4.2% error reduction.

Fine-Grained Air Pollution Inference with Mobile Sensing Systems: A Weather-Related Deep Autoencoder Model

Air pollution monitoring is a concerned issue in urban management. Nowadays, vehicle-based mobile sensing systems are deployed to enhance sensing granularity. However, in these systems, the number of online sensors may change over time and the long-term maintenance is costly. Therefore, an inference algorithm is necessary to maintain the high spatiotemporal granularity of pollution field under both dense and sparse sampling. In this paper, we propose an algorithm based on super-resolution scheme. To address the challenges of complex external factors and spatiotemporal dependencies, three modules are included: a heterogeneous data fusion subnet to extract useful information from external data, a spatiotemporal residual subnet to capture the spatiotemporal dependencies in pollution field, and an upsampling subnet to generate the fine-grained pollution map. Experiments on real-world datasets show that our model outperforms the state-of-the-art baselines.

Inferring fine-grained air pollution map via a spatiotemporal super-resolution scheme

Air pollution in a city is the major environmental risk to health. Mobile sensing has become a popular solution in recent years. However, it still suffers from problems such as lack of data and high system uncertainty. This is because that the data amount and distribution vary over time. To address the problems, this paper combines two classic data driven models -- Kriging and Inverse Distance Weighting (IDW). We adopt the Random Forest Algorithm (RF) to adaptively choose the more accurate models (Kriging or IDW) according to the features we extracted. The experiment based on real world testbed shows our adaptive method achieves up to 30.6% error reduction.

A Hybrid Air Pollution Reconstruction by Adaptive Interpolation Method

Air pollution is a global health threat. Nowadays, with the increasing amount of air pollution monitoring data from either conventional official stations or mobile sensing systems, the role of deep learning methods in pollution map recovery becomes gradually apparent. To address the challenges including the irregular sampling from mobile sensing and the non-interpretability of deep learning models, we proposed a deep autoencoder framework based inference algorithm. By separating the process of pollution generation and data sampling, this framework is able to deal with sampling under any irregular intervals in time and space. Also, we adopt a convolutional long short-term memory (ConvLSTM) structure to model the pollution generation after revealing its internal connections with an atmospheric dispersion model. Our algorithm is evaluated over a three-month real-world data collection in Tianjin, China. Results show our method can obtain up to 2x performance improvement over existing methods, benefited from its high robustness against different background pollution level and accidental sensor errors.

Yue Wang

Papers

Guiding the Data Learning Process with Physical Model in Air Pollution Inference

Fine-Grained Air Pollution Inference with Mobile Sensing Systems: A Weather-Related Deep Autoencoder Model

Inferring fine-grained air pollution map via a spatiotemporal super-resolution scheme

A Hybrid Air Pollution Reconstruction by Adaptive Interpolation Method

A deep autoencoder model for pollution map recovery with mobile sensing networks