Open AccessAnalysis of distribution characteristics of PM2.5 and health risk appraisal in northeast china through the geographically weighted regression model
Author(s) -
Zhe Zhu,
Yanting Zhang,
Xi Wang,
David Yong
Publication year - 2021
Publication title -
work
Language(s) - English
Resource type - Journals
eISSN - 1875-9270
pISSN - 1051-9815
DOI - 10.3233/wor-205373
Subject(s) - geographically weighted regression , statistics , regression analysis , bayesian multivariate linear regression , mean squared error , linear regression , logistic regression , multivariate statistics , econometrics , geography , mathematics
BACKGROUND: Recently, the frequent occurrence of air pollution greatly affects people’s health. OBJECTIVE: It aims to explore the spatial non-stationarity of PM2.5. METHODS: Geographically Weighted Regression (GWR) model is applied to fit spatial structure to linear regression model. The basic principles of GWR model are introduced. Bandwidth of weight function is optimized. Then, health risk of residents in northeast China is appraised according to PM2.5 distribution characteristics. A model for phycological health risk appraisal is established. RESULTS: From the linear analysis between the fitting results by designed GWR model and ground observation, the determination coefficient is 0.7, Relative Accuracy (RA) is 0.62, Mean Prediction Error (MPE) is 25.3, and Root Mean Squared Error (RMSE) is 26.2. The model prediction results are superior to those of ordinary multivariate regression model. Moreover, the risk of respiratory and cardiovascular diseases in the elderly is positively correlated with PM2.5 exposure. The illness of the elderly is related to weather factors such as average temperature, air pressure, and relative humidity. CONCLUSIONS: The study provides a reference for the research on risk of illness of people in PM2.5 environment.