Application of Daily Air Pollutant Index Forecasting Model Based on Semi-Empirical Statistical Theory:Case Study in Hanoi, Vietnam

摘要

This article illustrates Pham Ngoc Ho's forecasting model applied for daily air pollutant index, including PMJ〇, CO, NO, and 〇3. The authors have tested the model by analyzing 24-hour-per-day continuous monitoring data in 2017 - 2018 from the Nguyen Van Cu pennanent monitoring station to forecast daily air pollution index in Hanoi. The results have demonstrated that our model forecasts the air pollution index with the efficiency of 75-95% and 85-98% for the case of respectively 1-hour, 8-hour and 24-hour average in a day. In comparison with this model, the following models are applied and cited: Hanna SR’s simple statistical model which is tested using the same monitoring data from Nguyen Van Cu permanent monitoring station, the interpolation/extrapolation model of the author Duong Ngoc Bach which used monitoring data in 2012 at Nguyen Van Cu permanent monitoring station and Pongpiachan and Paovva's model of pollutants interacting with meteorological factors applied in Chiang-Mai, Thailand in 2015. The results of these 3 models have shown that, in contrast of Hanna SR's model that has relatively low accuracy, the remaining models have high accuracy. However, the model we use has an outstanding advantage of forecasting the air pollution index according to the daily forecast of meteorological factors on the mass media. This is a new approach that has never been reportedly applied to any contemporary modelling. This is the goal of the study.