Comparison of Nitrogen Dioxide Predictions During a Pandemic and Non-pandemic Scenario in the City of Madrid using a Convolutional LSTM Network

Traditionally, machine learning technologies with the methods and capabilities available, combined with a geospatial dimension, can perform predictive analyzes of air quality with greater accuracy. However, air pollution is influenced by many external factors, one of which has recently been caused by the restrictions applied to curb the relentless advance of COVID-19. These sudden changes in air quality levels can negatively influence current forecasting models. This work compares air pollution forecasts during a pandemic and non-pandemic period under the same conditions. The ConvLSTM algorithm was applied to predict the concentration of nitrogen dioxide using data from the air quality and meteorological stations in Madrid. The proposed model was applied for two scenarios: pandemic (January–June 2020) and non-pandemic (January–June 2019), each with sub-scenarios based on time granularity (1-h, 12-h, 24-h and 48-h) and combination of features. The Root Mean Square Error was taken as the estimation metric, and the results showed that the proposed method outperformed a reference model, and the feature selection technique significantly improved the overall accuracy.

Spatiotemporal neural network for estimating surface NO2 concentrations over north China and their human health impact.

Atmospheric nitrogen dioxide (NO2) is an important reactive gas pollutant harmful to human health. The spatiotemporal coverage provided by traditional NO2 monitoring methods is insufficient, especially in the suburban and rural areas of north China, which have a high population density and experience severe air pollution. In this study, we implemented a spatiotemporal neural network (STNN) model to estimate surface NO2 from multiple sources of information, which included satellite and in situ measurements as well as meteorological and geographical data. The STNN predicted NO2 with high accuracy, with a coefficient of determination (R2) of 0.89 and a root mean squared error of 5.8 µg/m3 for sample-based 10-fold cross-validation. Based on the surface NO2 concentration determined by the STNN, we analyzed the spatial distribution and temporal trends of NO2 pollution in north China. We found substantial drops in surface NO2 concentrations ranging between 9.1% and 33.2% for large cities during the 2020 COVID-19 lockdown when compared to those in 2019. Moreover, we estimated the all-cause deaths attributed to NO2 exposure at a high spatial resolution of about 1 km, with totals of 6082, 4200, and 18,210 for Beijing, Tianjin, and Hebei Provinces in 2020, respectively. We observed remarkable regional differences in the health impacts due to NO2 among urban, suburban, and rural areas. Generally, the STNN model could incorporate spatiotemporal neighboring information and infer surface NO2 concentration with full coverage and high accuracy. Compared with machine learning regression techniques, STNN can effectively avoid model overfitting and simultaneously consider both spatial and temporal correlations of input variables using deep convolutional networks with residual blocks. The use of the proposed STNN model, as well as the surface NO2 dataset, can benefit air quality monitoring, forecasting, and health burden assessments.

An air quality prediction model based on deep learning and wavelet analysis considering the COVID-19 pandemic factors

Based on the air pollution data in China from January 1,2014 to December 31,2020, the characteristics of extreme value and period of air quality in different regions on different time scales were studied by using wavelet analysis. Wavelet coherence analysis was used to evaluate the relationship between air quality and meteorological factors in the period of COVID-19. We found that the spatial characteristics of air quality changed significantly in summer. Generally, air pollution is more severe in spring and winter. During the lockdown period, the overall air quality in the study area improved significantly. In general, except for 03, the concentration of all other pollutants has dropped considerably. The improvement in air quality is a direct result of emission reductions due to the implementation of the COVID-19 blockade, which is unsustainable in the long term. Eventually, a prediction model attention_CNN_LSTM based on deep learning method is proposed in this paper. The experimental results show that the attention proposed in this study the model has a good prediction effect in the long-term prediction of air quality, but the attention mechanism's impact is lower. After shortening the prediction period, the attention_CNN_LSTM model has good prediction performance on most data sets, with average MAPE = 2.67% and RMSE = 2.29.

Deep Learning and Augmented Reality for IoT-based Air Quality Monitoring and Prediction System

During the pandemic of Corona-virus Disease 2019 (COVID-19), the whole world was confronted by a particularly high death toll and infection rate. Research has shown that air pollution plays a considerable part in the spread of certain illnesses and diseases. In the case of the COVID-19 pandemic, research has shown that increased air pollution has a negative effect on people's well-being and plays a role in the quick spread of the disease. Air pollution by itself affects the respiratory system of individuals which is aggravated, in addition, by a COVID19 infection. Some efforts have been made to use emerging technologies to combat the virus and its subsequent aerosol aspects to reduce transmission. In this context, we present an IoT system for Air Quality (AQ) monitoring and prediction using deep learning for data analysis and Augmented Reality (AR) for data visualization. The proposed system shows great potential for using Recurrent Neural Networks (RNN) and Long Short-Term Memory (LSTM) units as a framework for leveraging knowledge from time-series data of AQ. Moreover, integrating AR visualization for the proposed IoT system enables intuitive interaction between users and IoT devices and further improves visualization of AQ data which effectively contributes to easily conducting a deeper analysis of data and makes faster decisions. © 2021 IEEE.