Online Program Home
My Program

Abstract Details

Activity Number: 121 - Handling Large Dimensionality, Skewness and Non-Stationarity Through Multi-Resolution Spatial Modeling
Type: Topic Contributed
Date/Time: Monday, July 29, 2019 : 8:30 AM to 10:20 AM
Sponsor: Section on Statistics and the Environment
Abstract #304915 Presentation
Title: Using the MRA Approximation to Integrate Multiple Data Sources on Temperature
Author(s): Colin Lewis-Beck* and Veronica J. Berrocal and Joon Jin Song
Companies: and University of Michigan and Baylor University
Keywords: MRA; data fusion ; Bayesian; temperature ; spatial ; latent variable

Accurate and reliable spatio-temporal data on temperature are important for assessing the impact of extreme temperatures on health outcomes. However, daily temperatures are typically available from incongruous data sources with different types of biases and limitations. Monitors located at airports provide consistent and accurate measurements of air temperature, but are spatially sparse. Citizen science data are typically available in more urban areas, but have lower temporal frequency than the NOAA monitors, and might be prone to measurement errors due to a lower reliability of the measuring devices. Remote sensing data has high spatial and temporal resolution but provides estimates of land surface temperature, not air temperature. In this paper we propose a spatio-temporal modeling framework to fuse these three sources of data together in order to estimate air temperature fields during the summer. We use splines to account for the temporal structure of air temperature fields, while to account for spatial dependence and handle the large-dimensionality of the fields, we use a Multi-Resolution Approximation (M-RA). We demonstrate our model on data from the Industrial Midwest.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2019 program