Author : Christian Pelayo
Release : 2020
Genre : Atmospheric sciences
Kind : eBook
Book Rating : /5 ( reviews)
Book Synopsis Modeling PM2.5 Speciation Concentrations Over California Using the MISRV23 Aerosol Product by : Christian Pelayo
Download or read book Modeling PM2.5 Speciation Concentrations Over California Using the MISRV23 Aerosol Product written by Christian Pelayo. This book was released on 2020. Available in PDF, EPUB and Kindle. Book excerpt: Public health organizations are interested in better understanding the spatiotemporal variability of fine particulate matter (PM2.5) to relate exposure levels to human health effects. High levels of PM2.5 have been attributed to higher rates of human respiratory disease and premature death. While the health effects of total PM2.5 have been extensively studied, the health effects of speciated PM2.5 are poorly understood. The Interagency Monitoring of Protected Visual Environments (IMPROVE) and Chemical Speciation Network (CSN) ground monitor sites are capable of measuring speciated PM2.5. These data sets lack the necessary spatial resolution to capture the complex behavior of speciated PM2.5 because of emission, transport, and mixing processes. Aerosol optical depth (AOD) retrievals by the Multi-angle Imaging SpectroRadiometer (MISR) instrument aboard Terra have previously been incorporated into models to predict ground level PM2.5 concentrations. The newly reprocessed MISRV23 data set has a 4.4 km2 resolution for the entire global MISR record, making it useful for epidemiology studies (Chau et al., 2020).In this study, fractional AOD, spatial, and temporal variable predictors from the MISRV23 data set are coupled with North American Regional Reanalysis (NARR) planetary boundary layer height (PBLH) and colocated with CSN and IMPROVE PM2.5 data into a generalized additive model (GAM) capable of predicting speciated PM2.5 concentrations of elemental carbon, organic carbon, nitrate, and sulfate over the entire state of California. Each model has an R2 of 0.553, 0.37, 0.487, and 0.607 respectively. Model performance summaries indicate that every predictor variable is significant and that the models are capable of incorporating additional parameters to further improve performance. These models are used to generate surface maps of annual averaged speciated PM2.5 concentrations for the state of California.
