Activity Number:
|
413
- Analyses of Environmental Data
|
Type:
|
Contributed
|
Date/Time:
|
Thursday, August 12, 2021 : 2:00 PM to 3:50 PM
|
Sponsor:
|
Section on Statistics and the Environment
|
Abstract #318877
|
|
Title:
|
Functional Data Fusion of PM2.5 Observations and Satellite AOD Measurements
|
Author(s):
|
Zhengyuan Zhu* and Yueying Wang and Lily Wang
|
Companies:
|
Iowa State University and Iowa State University and Iowa State University
|
Keywords:
|
Kriging;
spatial statistics;
remote sensing;
non-parametric
|
Abstract:
|
Monitoring and forecasting PM2.5 is important for countries where air pollution is a serious public health issue. Current PM2.5 forecasts are mostly based on observations from monitoring stations, which have high temporal frequency but sparse and uneven spatial distribution. Aerosol Optical Depth (AOD) data from satellites such as MODIS has better spatial coverage but low temporal frequency. The fusion of PM2.5 data from stations and AOD data from satellites to provide hourly high-resolution PM2.5 data is useful for forecasting and epidemiology study. In this paper, we propose a novel data fusion framework using spatial functional data analysis tools. Efficient algorithms are developed to estimate the non-stationary mean and covariance structure using Bivariate spline and PACE. Estimates from the AOD data are then used to improve the spatial prediction of the PM2.5. Asymptotic results for the estimation and prediction are established, and exact simultaneous confidence bands (SCB) are developed. The proposed method is applied to data in the Beijing area in China, and our proposed approach outperforms several existing data fusion methods.
|
Authors who are presenting talks have a * after their name.