Deep-AIR: A Hybrid CNN-LSTM Framework for Air Quality Modeling in Metropolitan Cities

Abstract

Air pollution has long been a serious environmental health challenge, especially in metropolitan cities, where air pollutant concentrations are exacerbated by the street canyon effect and high building density. Whilst accurately monitoring and forecasting air pollution are highly crucial, existing data-driven models fail to fully address the complex interaction between air pollution and urban dynamics. Our Deep-AIR, a novel hybrid deep learning framework that combines a convolutional neural network with a long short-term memory network, aims to address this gap to provide fine-grained city-wide air pollution estimation and station-wide forecast. Our proposed framework creates 1x1 convolution layers to strengthen the learning of cross-feature spatial interaction between air pollution and important urban dynamic features, particularly road density, building density/height, and street canyon effect. Using Hong Kong and Beijing as case studies, Deep-AIR achieves a higher accuracy than our baseline models. Our model attains an accuracy of 67.6%, 77.2%, and 66.1% in fine-grained hourly estimation, 1-hr, and 24-hr air pollution forecast for Hong Kong, and an accuracy of 65.0%, 75.3%, and 63.5% for Beijing. Our saliency analysis has revealed that for Hong Kong, street canyon and road density are the best estimators for NO 2 , while meteorology is the best estimator for PM 2.5 .

Graph Neural Networks in IoT: A Survey

“…Air quality in our living environment is critical to our health especially during the Covid-19 pandemic. Deep-Air [95] introduced a novel hybrid CNN with LSTM to provide fine-grained city-wide air pollution estimation and station-wide forecasts on air quality. In smart industry, Wen et al [279] developed a hybrid sensor fusion system to accurately predict the remaining useful life (RUL) of IoT-enabled complex industrial systems.…”

“…LSTMs are a type of recurrent neural network (RNN) algorithm that boosts the forgetfulness problem in sequential inputs. For example, Hasan et al presented a CNN-LSTM based approach for electricity theft detection in smart grid systems [98]; Han et al designed a hybrid CNN-LSTM framework for air quality modeling in metropolitan cities [95]; while Yang et al applied a parallel convolutional RNN for emotion recognition using multi-channel EEG data [298].…”