With increasing air pollution and its harmful effect on the residents of developing countries, the prediction and analysis of pollutants have become an important research aspect. This study focuses on the spatiotemporal prediction of hourly particulate matter with different deep-learning modeling techniques for Delhi, India. The secondary data of particulate matter concentrations and the meteorological parameters for the four static monitors in the area are collected from Central Pollution Control Board (CPCB) for dates between January 2019 and April 2021. The study area of South Delhi is divided into hexagonal grids. The data sets at the centroid of each grid are formulated with the spatial interpolation method of inverse distance weighting and Kriging. The hexagonal grids are required to collate the data coming from dynamic monitors. Three models with convolutional neural network (CNN), long short-term memory (LSTM), and CNN-LSTM are developed for a total of 15 cells. To evaluate developed models, mean absolute error and root mean square error are used. The results from prediction models show that CNN-LSTM models outperform the other two models. The predictions are accurate for the CNN-LSTM model compared with the values obtained from the static monitor. Also, compared with the existing and individual models, the proposed hybrid CNN-LSTM model performed better for most of the cells. The prediction models can also provide the pollutant concentration on various routes, which can assist residents in making travel choices based on the air pollution prediction information. Planners and practitioners can replicate the developed models in other regions.


    Access

    Download

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Deep-Learning Spatiotemporal Prediction Framework for Particulate Matter under Dynamic Monitoring


    Additional title:

    Transportation Research Record


    Contributors:


    Publication date :

    2022-03-03




    Type of media :

    Article (Journal)


    Type of material :

    Electronic Resource


    Language :

    English



    Deep hybrid learning framework for spatiotemporal crash prediction using big traffic data

    Mohammad Tamim Kashifi / Mohammed Al-Turki / Abdul Wakil Sharify | DOAJ | 2023

    Free access


    A Deep Spatiotemporal Trajectory Representation Learning Framework for Clustering

    Wang, Chao / Huang, Jiahui / Wang, Yongheng et al. | IEEE | 2024


    Spatiotemporal Deep-Learning Networks for Shared-Parking Demand Prediction

    Liu, Yonghong / Liu, Chunyu / Luo, Xia | ASCE | 2021