In recent years, the emission problems associated with freight transportation in urban areas have become increasingly severe, significantly affecting residents' health and the environment. To address this, this study focuses on the Chengdu-Chongqing urban cluster as a typical example and proposes a pollutant measurement and prediction method based on forecasting highway freight transportation volume in the region. The study utilizes OD (Origin-Destination) data of road freight from the online freight exchange platform (OFEP) in the Chengdu-Chongqing area, collected between November 1, 2017, and March 7, 2018. Taking into account the time and spatial characteristics of freight data, a spatiotemporal prediction model of OD freight volume is developed using ConvLSTM. For comparison, an LSTM model is also established. The results indicate that the ConvLSTM model, which incorporates spatial characteristics, achieves higher prediction accuracy. Using the predicted freight demand data, carbon emissions are calculated using a top-down approach, revealing the distribution of carbon emissions across each OD pair in the Chengdu-Chongqing urban cluster. The analysis of short-term emission trends provides valuable insights into the precise regulation of carbon emissions from road freight.

Meitong An, Yuhong Fu. Forecasting Road Freight Demand and Estimating Carbon Emissions Using ConvLSTM: The Chengdu-Chongqing Urban Cluster Case. Academic Journal of Computing & Information Science (2025), Vol. 8, Issue 1: 71-79. https://doi.org/10.25236/AJCIS.2025.080110.

[1] Deloison, T., Hannon, E., Huber, A., Heid, B., Klink, C., Sahay, R., & Wolff, C. (2020). The future of the last-mile ecosystem: Transition roadmaps for public-and private-sector players. World Economic Forum.

[2] An, R., Hua, G., & Dong, N. (2015). Freight volume forecasting in Nanning based on BP neural network. Transportation Research, 1(1), 58-64.

[3] Xu, J., Gatpandan, P. H., Gatpandan, M. P., & Gao, Z. (2021, July). Freight and Passenger Volume Prediction using Multiple Regression. In 2021 International Conference on Machine Learning and Intelligent Systems Engineering (MLISE) (pp. 77-83). IEEE.

[4] Wang, Z., Huo, Y., & Liu, Z. (2018). Prediction Model on Energy Consumption of Highway Transportation in Inner Mongolia Based on ARMA. In Green Intelligent Transportation Systems: Proceedings of the 7th International Conference on Green Intelligent Transportation System and Safety 7 (pp. 105-114). Springer Singapore.

[5] Wu, Y., Wang, S., Zhang, Y., & Zhang, J. (2021, June). A BP neural network model for the demand forecasting of road freight transportation system. In 2021 International Symposium on Computer Technology and Information Science (ISCTIS) (pp. 264-268). IEEE.

[6] Lin, N., Yuan, Y., & Li, H. (2023). Research on logistics volume prediction model based on ARIMA-LSTM. Highlights in Business, Economics and Management, 18, 98-104.

[7] Song, Y., & Lu, Z. (2024). Logistics sorting center volume prediction based on time series prediction LSTM model. Frontiers in Traffic and Transportation Engineering, 4, 115-122.

[8] Wang, H., Zhang, L., & Li, J. (2023). A hybrid deep learning model for freight volume forecasting based on GRU and attention mechanism. Computational Intelligence and Neuroscience, 2023, 1-12.

[9] Ma, X., Zhang, J., Du, B., Ding, C., & Sun, L. (2018). Parallel architecture of convolutional bi-directional LSTM neural networks for network-wide metro ridership prediction. IEEE Transactions on Intelligent Transportation Systems, 20(6), 2278-2288.

[10] Liu, J., Ou, G., & Guo, Z. (2018, December). Regional Freight Volume Forecasting with Incomplete Data of Origin/Destination Freight Volumes. In 2018 IEEE International Conference on Industrial Engineering and Engineering Management (IEEM) (pp. 839-843). IEEE.

[11] Xiangfeng, W. (2022). Prediction and Analysis of Freight Transportation Demand Based on GM (1, 1)-Markov Model. Academic Journal of Business & Management, 4(4), 48-51.

[12] Cohen, J. B., and Wang, C. (2014), Estimating global black carbon emissions using a top-down Kalman Filter approach, J. Geophys. Res. Atmos., 119, 307– 323.

[13] Wiebe, K., Gandy, S. and Lutz, C. (2016) Policies and Consumption-Based Carbon Emissions from a Top-Down and a Bottom-Up Perspective. Low Carbon Economy, 7, 21-35.

[14] Gurney, K. R., Liang, J., O'Keeffe, D., Patarasuk, R., Hutchins, M., Huang, J., Rao, P., & Song, Y. (2019). Comparison of Global Downscaled Versus Bottom-Up Fossil Fuel CO2 Emissions at the Urban Scale in Four U.S. Urban Areas. Journal of Geophysical Research D: Atmospheres, 124(5), 2823-2840.

[15] Li, L. (2019). Structure and influencing factors of CO 2 emissions from transport sector in three major metropolitan regions of China: Estimation and decomposition. Transportation, 46, 1245-1269.

[16] Shi, X., Chen, Z., Wang, H., Yeung, D. Y., Wong, W. K., & Woo, W. C. (2015). Convolutional LSTM network: A machine learning approach for precipitation nowcasting. Advances in neural information processing systems, 28.

[17] IPCC. (2023). Climate Change 2023: Synthesis report. Contribution of working groups I, II and III to the sixth assessment report of the Intergovernmental Panel on Climate Change [Core Writing Team, A. S. S. de Lima, & R. K. Pachauri (Eds.)]. IPCC, Geneva, Switzerland.