Search

Satellite-derived measurements are negatively impacted by cloud cover and surface reflectivity. These biases must be discarded and significantly increase the amount of missing data within remote sensing images. This paper expands the application of a partial convolutional neural network (PCNN) to incorporate depthwise convolution layers, conferring temporal dimensionality to the imputation process. The addition of a temporal dimension to the imputation process adds a state of successive existence within the dataset, which spatial imputation cannot capture. The depthwise convolution process enables the PCNN to independently convolve the data for each channel. The deep learning system is trained with the Community Multiscale Air Quality model-simulated tropospheric column density of Nitrogen Dioxide (TCDNO2) to impute TROPOspheric Monitoring Instrument TCDNO2. The depthwise PCNN model achieves an index of agreement of 0.88 and outperforms the default PCNN models, with and without temporal dimensionality of data, and conventional data imputation methods such as inverse distance weighting by 4–7 and 10–15% in the index of agreement and correlation, respectively. The model demonstrates more consistency in the reconstruction of TROPOspheric Monitoring Instrument tropospheric column density of NO2 images. The model has also demonstrated the accurate imputation of remote sensing images with over 95% of the data missing. PCNN enables the accurate imputation of remote sensing data with large regions of missing data and will benefit future researchers conducting data assimilation for numerical models, emission studies, and human health impact analyses from air pollution. [ABSTRACT FROM AUTHOR]

This study proposes a novel two‐step deep learning (DL) model for estimating surface NO2 concentrations using satellite data over the contiguous United States (CONUS) from 2005 to 2019. The first phase of the model uses partial convolutional neural network (PCNN), an advanced DL model that accurately imputes gaps between surface NO2 stations and creates 5,478 daily‐mean NO2 grids (PCNN‐NO2) of the 2005–2019 period over the study area. We then feed the PCNN‐NO2, along with other predictor variables, into a deep neural network (DNN) to estimate surface NO2 levels, achieving exceptional performance with a correlation coefficient of 0.975–0.978, a mean absolute bias of 0.99–1.38 ppb, and a root mean square error of 1.47–1.97 ppb. Spatial cross‐validation results also indicate strong spatial performance of PCNN‐DNN surface NO2 estimates. In addition to its accurate estimates, the PCNN‐DNN model consistently generates estimated NO2 grids without any missing values, improving the quality of various applications such as emission reduction strategies and public health studies. Between 2005 and 2019, the 5,478 daily estimated NO2 grids over the CONUS reveal significant reductions in NO2 levels in 14 major urban environments: Washington D.C. (−43%), New York (−45%), Los Angeles (−38%), Chicago (−25%), Boston (−43%), Houston (−34%), Dallas (−40%), Philadelphia (−41%), Phoenix (−38%), Detroit (−20%), Denver (−23%), Atlanta (−0.7%), Cincinnati (−38%), and Pittsburgh (−56%). Furthermore, the study shows that the denser urban regions that in‐situ stations are installed in, the higher the difference between in‐situ observations and regional‐mean NO2 levels. Plain Language Summary: Despite the significance of NO2 monitoring stations, their number and spatial coverage is limited, leaving many regions with no NO2 observations. This limitation calls for the development of advanced models to fill the gaps between NO2 stations and create gap‐free spatiotemporally‐consistent grids of surface NO2 concentrations. Recent advances in deep learning (DL) algorithms have enabled them to fill the gaps in various images, such as those representing the human face, nature, etc. with plausible contents. This technology can also be applied to fill the gaps between surface NO2 stations when the model is trained on datasets such as satellite tropospheric column density of NO2 images, available in almost all regions. In this study, we introduce a novel DL model to construct daily gap‐free grids of estimated NO2 levels over the contiguous United States from 2005 to 2019, obtaining exceptional performance with a correlation coefficient of 0.975–0.978. The first phase of the model uses a partial convolutional neural network, trained mainly by satellite images, to fill the gaps between surface NO2 stations, and the second phase utilizes a deep neural network to bias correct the output of the first phase. Key Points: A novel two‐step deep learning model is used to estimate surface NO2 levels, with exceptional accuracy, over the contiguous United StatesThe partial convolutional neural network used satellite images to fill the gaps between surface NO2 stationsNO2 levels decreased significantly in most urban environments in the United States from 2005 to 2019 [ABSTRACT FROM AUTHOR]

The limited number of nitrogen dioxide (NO2) surface measurements calls for the development of highly accurate approaches to estimating surface NO2 concentrations. In this study, we leverage a new satellite instrument, the TROPOspheric Monitoring Instrument (TROPOMI), along with other predictor variables, to estimate daily surface NO2 concentrations over Texas in 2019. We use the deep convolutional neural network (Deep‐CNN), an advanced deep learning algorithm, to obtain estimates and achieve a correlation coefficient (R) of 0.91, an index of agreement (IOA) of 0.95, and a mean absolute bias (MAB) of 1.75 ppb in surface NO2 estimation. Additionally, we leverage a novel approach, shapley additive explanations (SHAP), to describe how Deep‐CNN understands each predictor variable. The SHAP results show that the Deep‐CNN model has an advanced understanding of the data set, revealing that TROPOMI closely captures levels of NO2. In addition, we show the superiority of our Deep‐CNN model at estimating surface NO2 over other well‐known machine learning and regression models in the field, including the support vector machines (SVM), random forest (RF), and multiple linear regression (MLR). Although SVM and RF show strong capabilities at estimating surface NO2 concentrations, their accuracy is inferior to that of the Deep‐CNN model, ranking second and third in model accuracy in this study. The MLR, however, shows a poor ability at NO2 estimation and ranks last among all models. Testing the impact of sample size on model performance, we also show that, compared to other models, Deep‐CNN needs more samples to trigger its strength at surface NO2 estimation. Plain Language Summary: While relatively low in concentrations, nitrogen dioxide (NO2) is an air pollutant that has severe health and environmental impacts. Although air quality ground stations can accurately measure NO2 concentrations, they have low spatial coverage owing to land accessibility, budget limitations, and other factors. Accurately estimating surface NO2 concentrations requires new approaches that address these spatial coverage limitations. Recent advances in satellite instruments have enabled the monitoring of various atmospheric constituents such as NO2 over entire regions. As a result, instead of using conventional means of estimating surface pollutants such as model simulations or statistical methods, scientists have developed deep learning (DL) algorithms to improve the accuracy of estimates. For the first time, this study leverages the deep convolutional neural network (Deep‐CNN) and the TROPOspheric Monitoring Instrument (TROPOMI), together with other environmental predictor variables, to produce accurate estimates of surface NO2 concentrations over Texas in 2019. One problem with DL models is the difficultly of interpreting their behavior concerning the data set. In response, we use a novel approach, the shapley additive explanations, to explain how the Deep‐CNN model understands the data set. We also compare the performance of our Deep‐CNN model with other well‐known models in the field. Moreover, we test the impact of sample size on performance of various models to better analyze capability of Deep‐CNN and other commonly‐used models for surface NO2 estimation. Key Points: Deep‐CNN is highly capable of estimating surface NO2 concentrations (CV‐R: 0.91)The shapley additive explanations approach shows that the deep convolutional neural network has an advanced understanding of the data setTROPOMI, a new accurate satellite instrument, plays an essential role in ground‐level NO2 estimation [ABSTRACT FROM AUTHOR]

This paper proposes a deep learning model that integrates a convolutional neural network and a gated recurrent unit with groups of neighboring stations to accurately predict PM2.5 concentrations at 25 stations in Seoul, South Korea. The deep learning model uses observations obtained from one Korea Meteorological Administration (KMA) station, 25 National Institute of Environmental Research (NIER) stations, and 28 automatic weather stations (AWSs) throughout Seoul. To train the deep learning model, we use all available meteorological and air quality data observed between 2015 and 2017. With the trained model, we predict PM2.5 concentrations at all 25 NIER stations in Seoul for 2018. This study also proposes a geographical polygon group model that determines the optimal number of neighboring NIER stations required to increase the accuracy of PM2.5 concentration predictions at the target station. Comparing the model measures for each of the 25 monitoring sites in 2018, we find that the geographical polygon group model achieves an index of agreement of 0.82–0.89 and a Pearson correlation coefficient of 0.70–0.83. Compared to the method using only meteorological and air quality data from one target station (average IOA = 0.77) to predict PM2.5 concentrations at the 25 stations in Seoul, the proposed method using geographical correlation-based neighboring NIER stations as polygonal groups (average IOA = 0.85) improves the PM2.5 prediction accuracy by an average of about 10%. This approach, based on deep learning, can be updated to predict air pollution or air quality indices up to several days in advance. [ABSTRACT FROM AUTHOR]