Nonlinear regression models to forecast PM$_{2.5}$ concentration in Wuhan, China

TL;DR

This paper develops nonlinear regression models for predicting PM$_{2.5}$ levels in Wuhan, achieving precise daily forecasts and effectively capturing high and low pollution days, with integration of climate forecasts to improve accuracy.

Contribution

It introduces a novel nonlinear regression approach for PM$_{2.5}$ forecasting, including both point and interval models, and combines them with climate system data for enhanced prediction accuracy.

Findings

Single-value forecast bias about 6 μg/m^3

Interval forecast covers 60%-80% observed samples

Model demonstrates good independent forecasting capacity

Abstract

Forecasting PM$_{2.5}$ concentration is important to solving air pollution problems in Wuhan. This paper proposes a PM$_{2.5}$ concentration forecast model based on nonlinear regression, including a single-value forecast model and an interval forecast model. The single-value forecast model can precisely forecast PM$_{2.5}$ concentration for the next day, with forecast bias about 6 $\mu g/m^3$ in goodness of fit analysis. The interval forecast model can efficiently forecast high-concentration and low-concentration days, which covers 60%-80% observed samples in model validation. Moreover, this paper combines the PM$_{2.5}$ concentration forecast model with NCEP Climate Forecast System Version 2 to realize its forecast application, then develops NCEP CFS2's PM$_{2.5}$ concentration forecast model to enhance forecast accuracy. The results indicate that the PM$_{2.5}$ concentration forecast model has good capacity for independent forecasting.