Applying Regression Analysis to the Study of Suspended Particulates

碩士 === 國立中興大學 === 水土保持學系所 === 101 === In recent years, due to the rapid economic development of Taiwan, development of land use, and natural disasters, the probability of dust events was increased. Studies have pointed out that the smaller particles in the air are more easily inhaled and deposited i...

Full description

Bibliographic Details
Main Authors: Yong-Jun Chen, 陳詠鈞
Other Authors: Chang-Hai Chien
Format: Others
Language:zh-TW
Published: 2013
Online Access:http://ndltd.ncl.edu.tw/handle/99912150715824204913
Description
Summary:碩士 === 國立中興大學 === 水土保持學系所 === 101 === In recent years, due to the rapid economic development of Taiwan, development of land use, and natural disasters, the probability of dust events was increased. Studies have pointed out that the smaller particles in the air are more easily inhaled and deposited in the lungs resulting in damage to affect human health. Therefore, to understand which factors are affecting the suspended particles PM10 and PM2.5 concentrations of the most important factors are very urgent and important issue nowadays. In this study, the data were collected from 6 air quality monitoring stations (i.e., Fengyuan, Lunbei, Nantou, Taixi, Xianxi, Xingang) lasted a total of 19 years (from 1993 to 2011). Using suspended particles PM10 concentration and PM2.5 concentration as response variable, weather factors (i.e., temperature, relative humidity, rainfall, and wind speed) and chemical factors (i.e., CO, NO2, O3, SO2, NO, and NOx) as explanatory variables for regression analasis. In this study, the data were collected from the air quality monitoring stations and arranged into daily average. Draw up maximum and minimum value, and use this basis to weed out the outliers and the unreasonable values, and arrange each factor’s statistical magnitude. Boxplots and bar charts of PM10 concentration were generated to facilitate the comparison as well as show the variations. Using the statistical software SPSS to correlation analysis. Investigate the correlation between the intensity factors and analyzed with the suspended particles. Finally, use the multiple linear regression analysis to all of data, and data of the wet season and the drought season. As a result, O3 concentration of suspended particles is the most important factor and the second one is CO. Other important factors are different due to the station category, or the wet season and the drought season. On the other hand, factors affecting the concentration of PM10 and PM2.5 concentrations are different.