|
Activity Number:
|
296
- Advances in Inference for Massive Spatio-Temporal Environmental Data with Applications in Remote Sensing
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Tuesday, July 31, 2018 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
ENAR
|
|
Abstract #329745
|
|
|
Title:
|
Dynamic Fused Gaussian Process for Massive Sea Surface Temperature Data from MODIS and AMSR-E Instruments
|
|
Author(s):
|
Emily L. Kang* and Pulong Ma
|
|
Companies:
|
University of Cincinnati and University of Cincinnati
|
|
Keywords:
|
Kalman filter;
Remote sensing;
Smoothing;
Spatio-temporal statistics
|
|
Abstract:
|
Sea surface temperature (SST) is a key climate and weather measurement, which plays a crucial role in understanding climate systems. SST datasets collected from satellite instruments are often noisy and incomplete in space. In addition, different instruments usually produce SST data products at different spatial resolutions. We propose a Dynamic Fused Gaussian Process (DFGP) model that enables fast statistical inference such as smoothing and filtering for massive datasets. The change-of-support problem is also explicitly addressed in DFGP when statistical inference is made based on different sources of data whose spatial resolutions are incompatible. We also develop a stochastic Expectation-Maximization (EM) algorithm to allow fast parameter estimation in a distributed computing environment. The proposed DFGP is applied to a total of 3.5 million data for SST in a one-week period in tropical Pacific Ocean area from the Moderate Resolution Imaging Spectroradiometer (MODIS) and the Advanced Microwave Scanning Radiometer-Earth Observing System (AMSR-E) instruments.
|
Authors who are presenting talks have a * after their name.
