Remote Sensing, Uncertainty Quantification and a Theory of Data Systems Workshop: February 12-14, 2018


This workshop was held at the Cahill Center, California Institute of Technology


This Workshop invited statisticians, applied mathematicians, computer scientists, experts in remote sensing technology, and Climate and Earth System scientists to convene to review, discuss, and plan research on issues related to large-scale, efficient analysis of distributed data using spatial statistical methods. Remote sensing data are natural inputs to spatial statistical algorithms, but in many cases data are massive, and are stored in different physical locations. These data must be brought together in some way in order to estimate spatial covariance functions, but moving data to a central location for analysis is tedious at best and impossible at worst. Some remote data reduction is almost certainly necessary, but how much? What are the consequences for inference? The fundamental issue underlying these questions is how to navigate the trade-space between computational and transmission costs versus uncertainty in the estimates or inferences that are ultimately produced. The Workshop was organized around the following themes:

  • Data systems and their architectures especially at NASA and NOAA
  • Multi-layer network models for data systems
  • The computational–statistical trade-off: theory and application
  • Spatial statistics with distributed data
  • Case study problems with uncertainty requirements and cost limitations

Schedule and Supporting Media

Confirmed speakers for this event were:

Poster Session
Talks and Abstracts

Monday, February 12, 2018
Cahill Center, California Institute of Technology

Description Speaker Slides
Opening Remarks Amy BravermanJet Propulsion Laboratory; Jessica Matthews, NCSU/NOAA
Welcome/SAMSI David Banks, SAMSI Director; Richard Smith, SAMSI Associate Director
Welcome/CD3 and CDST George Djorgovski, Caltech; Dan Crichton, Jet Propulsion Laboratory
Distributed Access and Analysis: NASA Mike Little, NASA
Satellites and Stovepipes Jay Morris, NOAA
The Statistical Computational Trade-off Venkat Chandrasekaran, Caltech
Approximate Likelihoods Richard Smith, UNC-CH/SAMSI
Data System Architectures Dan Crichton, Jet Propulsion Laboratory
The ToDS Problem Maggie Johnson, SAMSI/NCSU
Multilayer Modeling and Analysis of Complex (Systems) Data Manlio De Domenico, Bruno Kaiser Foundation
Optimization Working Group Jessica Matthews, NOAA
Emulators Working Group Emily Kang, University of Cinncinati
Spatial Retrieval Working Group Jon Hobbs, Jet Propulsion Laboratory
Discussion Bruno Sanso, University of California, Santa Cruz (UCSC); Ansu Chatterjee, University of Minnesota; David Banks, SAMSI/Duke
Poster Session and Reception

Tuesday, February 13, 2018
Cahill Center, California Institute of Technology

Description Speaker Slides
Multi-resolution Approaches for Big Spatial Data Matthias Katzfuss, Texas A&M
DISK: A Divide and Conquer Bayesian Approach to Large Scale Kriging Rajarshi Guhaniyogi, UCSC
Optimization for Distributed Data Systems: An Overview and Some Theoretical Results Zhengyuan Zhu, Iowa State
High Performance Computing and Spatial Statistics: an overview of recent work at NCAR Dorit Hammerling, NCAR
The Earth System Grid Federation as a Testbed for Global, Distributed Data Analytics Luca Cinquini, Jet Propulsion Laboratory
Environmental Exposure in Environmental Epidemiological Studies: modeling approaches and challenges Veronica Berrocal , University of Michigan
Climate Science Hui Su, Jet Propulsion Laboratory
Sea-Ice Modeling and Analysis Carmen Boening, Jet Propulsion Laboratory Presentation NOT Available
Carbon Cycle Science Vineet Yadav, Jet Propulsion Laboratory
Discussion: Agenda for ToDS Research
Wrap-up, Plans for Wednesday

Wednesday, February 14, 2018
Cahill Center, California Institute of Technology

Description Speaker Slides
Discussion and Planning

Questions: email