Augmentation and Integration of Measurements and Simulation Data in Heterogeneous Environments
Abstract
Augmentation and Integration of Measurements and Simulation Data in Heterogeneous Environments. Discovery is often guided by observations from experimental measurements and from numerical simulations. As a result, the extent to which we can advance scientific discovery is directly related to our ability to accurately measure, or detect, an object or event of interest. This basic requirement cuts across scientific disciplines and application areas, from resolving nanoscale processes to measuring phenomena that take place on geophysical and even astrophysical scales. Performing accurate measurements is often, however, vey challenging due to sensor resolution. The interpretation of the measurements is at times equally or more difficult due to uncertainties in the sensor data, and due to heterogeneities in the environment where the measurements are performed. In this proposal, we establish an accurate and rigorous framework to address these basic challenges that impede discovery. Our approach will rely on combining measurements and simulation data. The specific challenge that will be used to guide our development of this framework is performing accurate measurements in heterogeneous turbulent environments. We will consider two important problems: First, we will examine how we can optimally augment limited measurement data provided at a relatively coarse resolution. Second, we will attempt to interpret sensor signals whose origin is from an unknown remote source, and which have been encoded by the stochasticity of the turbulent environment. We will focus on scalar measurements and develop optimal techniques to interpret, or decode, them. The same techniques can later be applied to acoustic sensors. In each of these problems, we will further increase the complexity by introducing uncertainty in the measurements and the environmental conditions. In order to systematically assess the influence of uncertainty, we will store multi-resolution representations of the flow fields of interest as part of the Johns Hopkins Turbulence Database (JHTDB, http://turbulence.pha.jhu.edu), and provide open user access to the entire data. We will also host our own augmented experimental data so they are available as a benchmark for the community. Therefore, in addition to enabling the proposed e?ort the database will also serve as computational laboratory for the wider scientific community to develop new data-augmentation algorithms using this platform.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Nov 23, 2016
- Source ID
- N000141612542
Entities
People
- Tamer A. Zaki
Organizations
- Johns Hopkins University
- Office of Naval Research
- United States Navy