Long-term forecasting of nitrogen dioxide ambient levels in metropolitan areas using the discrete-time Markov model

Long-term forecasting of nitrogen dioxide ambient levels in metropolitan areas using the discrete-time Markov model

A. Nebenzal & B. Fishbain

Abstract

Air pollution management and control are key factors in maintaining sustainable societies. Air quality forecasting may significantly advance these tasks. While short-term forecasting, a few days into the future, is a well-established research domain, there is no method for long-term forecasting (e.g., the pollution level distribution in the upcoming months or years). This paper introduces and defines long-term air pollution forecasting, where long-term refers to estimating pollution levels in the next few months or years. A Discrete-Time-Markov-based model for forecasting ambient nitrogen oxides patterns is presented. The model accurately forecasts overall pollution level distributions, and the expectancy for tomorrow’s pollution level given today’s level, based on longitudinal historical data. It thus characterizes the temporal behavior of pollution. The model was applied to five distinctive regions in Israel and Australia and was compared against several forecasting methods and was shown to provide better results with a relatively lower total error rate.

Code

Please find the MATLAB package here. The main file is the MarkovModelingMain.m. Running the main file you should choose between three types of models: Threshold exceeding model, Transition matrix and 2nd order markov model (see the manuscript for details).
In case of choosing the transition matrix the application also creates the stationary distribution matrix matrix.

The zip file contains both sample data with and without wind info. The main files contain detailed info how to run the files.

Please cite the following paper in any future publication using this packageA. Nebenzal and B. Fishbain, “Long Term Forecasting of Nitrogen Dioxide in Metropolitan Areas Using the Discrete Time Markov Model”, Environmental Modelling and Software, 107:175-185, 2018.