Spatial data has spatial heterogeneity when assessing site-specific concentrations the temporal and spatial variations in air pollutant concentrations may be higher in different weather conditions and areas where the land surface is uneven Since the existing spatial interpolation methods still have uncertainties the estimated concentrations are biased The purpose of this study is to reduce the uncertainty of previous estimates Combined with EPA and Airbox data an smart air pollution exposure estimation model was established to estimate exposure to PM2 5 concentrations using a representative EPA monitoring station The Modeling of site-specific exposures uses inverse distance weighted interpolation between data from a set of representative air quality stations which are generated through spatial clustering analysis by large amounts of data from low-cost sensors This study uses leave-one-out cross-validation to verify the feasibility of the model And compared with Kriging method and inverse distance weighting method We found that the developed method can generate a better exposure database by selecting suitable sites for spatial interpolation smartly with considering clustering of air quality regions that are differentiated by local weather and terrain conditions compared with traditional spatial interpolation methods kriging and inverse distance weighting The developed exposure database will support further analysis of the air pollutants on related health effects
Date of Award | 2019 |
---|
Original language | English |
---|
Supervisor | Pi-cheng chen (Supervisor) |
---|
Exposure Assessment of Fine Particulate Matters with Smart Spatial Interpolation Based on Low-Cost Sensors
昱廷, 林. (Author). 2019
Student thesis: Doctoral Thesis