EcoSta 2019: Start Registration
View Submission - EcoSta2019
A0773
Title: Spatial modeling of ground-level PM2.5 in Taiwan based on two types of data Authors:  Chi-Wei Lai - Academia Sinica (Taiwan) [presenting]
Hsin-Cheng Huang - Academia Sinica (Taiwan)
ShengLi Tzeng - National Sun Yat-sen University (Taiwan)
Abstract: There are two systems to monitor fine particulate matter (PM2.5) in Taiwan. One consists of 77 monitoring stations of the Environmental Protection Administration, which provides high-quality measurements. The other one involves a large number of low-cost internet-of-things devices called AirBoxes, which produce less precise measurements but with much broader coverage. We propose a spatial model to obtain spatial prediction at any location in Taiwan by combining these two types of data. In addition, we develop a Shiny application that automatically identifies unusual measurements and shows the current PM2.5 concentration map with uncertainty quantification based on the proposed method.