Most of our work involves Radar Remote Sensing in its various forms, with a major specialization in Radar Interferometry. We address both the development and implementation of radar systems and their application to scientific studies of the Earth and planets.
Student Research Pages
Albert Chen: InSAR Studies of Greenland Ice
Lauren Wye: Scattering from Saturn's Moon Titan and Other Satellites
Ann Chen: Phase Signatures of Earth's Ionosphere
Hrefna Gunnarsdottir: Bistatic Scattering from Mars' Surface
Piyush Shankar Agram: Persistent Scattering and Aseismic Crustal Deformation
Cody Wortham: Airborne Measurement of Deformation
Some of our other ongoing areas of research are:
- Crustal Deformation from Earthquakes
Our emphasis here is on i) detecting inter-seismic and potentially pre-seismic transient strains, which remain elusive and raise a major challenge to our understanding of the earth, and ii) modeling faulting and crustal rheology from vector co- and post-seismic displacement maps, which complement conventional seismological and geodetic measurements. Sjonni Jonsson has analyzed this radar interferogram of the Hector Mine earthquake
- Volcanological Studies
Deformation is also important in understanding volcanic processes. Our work here includes i) modeling magma migration from the spatial and temporal extent of deformation, ii) quantifying pressure changes at depth resulting from magma intrusion beneath volcanic edifices, and iii) analyzing the spatial extent of new material deposited during eruptions. Sjonni also has produced and modeled deformation at Sierra Negra in the Galapagos Islands.
- Earth's polar regions and the world's climate
Ice sheets and glaciers in the Earth's polar regions both control and reflect changes in the world climate. InSAR allows us to investigate the mass balance, or the accumulation and loss of ice, in these remote areas. Weber Hoen's thesis examines how InSAR correlation measurements help us estimate ice accumulation rates worldwide. Here is Shadi Oveisgharan's accumulation map of part of Greenland, and we need to apply this method to the full Antarctic continent to understand global warming or cooling.
- Other studies of ground subsidence
Our group is pioneering other new applications for InSAR. We have found these data to be useful for studying many geophysical phenomena of strong scientific value and societal benefit, such as the study and management of groundwater aquifer systems. Joern Hoffmann has concentrated on the use of InSAR to better model the flow of water underground (figure here) . Other fields under investigation include landslides, floods, oil extraction, and coastal erosion.
- Mapping the surfaces of planets and moons in our Solar System
Lauren Wye has developed data processing and analysis tools to study radar scattering from the surface of Titan, Saturn's largest moon, using the radar system on board the NASA Cassini spacecraft. She calibrated the instrument and subsequently derived surface dielectric constant and roughness estimates from radar scattering models.
New technique development
- Persistent scatterer analysis
InSAR analyses are often limited by decorrelation in regions containing vegetation or other time-variable surfaces. We have extended and improved the technique of persistent scatterers to be of use in natural terrains, so that interferograms may be produced in areas previously resistant to InSAR. Andy Hooper produced this image of deformation in Long Valley Caldera using his improved method.
- ScanSAR interferometry
Most radar satellites are in orbits with monthly repeat cycles, which limits the number of times a region may be visited by an orbiting SAR and also makes it difficult to observe rapidly occurring deformations. Ana Bertran-Ortiz has developed a method to use InSAR with the ScanSAR mode of modern satellites to achieve weekly revisits.
- Long-code waveform design
Planetary radar systems such as Arecibo and Goldstone must use very long waveforms to provide enough coherent integration to overcome the very weak received radar echoes. Leif Harcke applied long-code methods to these radars to overcome the problems of interfering ambiguities and "clutter" that can obscure faint radar signals, as shown in this image of ice at the north pole of Mercury.
- Split-beam processing
Another limitation of conventional InSAR is that only the line-of-sight component of deformation is measured on a single. Noa Bechor has implemented a method to obtain 2-D displacements from a single interferogram, so that a pair of ascending/descending interferograms now yields the 3-D deformation field. This greatly enhances our ability to model the surface.
Signal processing, design, and analysis
- Phase Unwrapping
Phase unwrapping is a multidimensional signal processing technique that permits unambiguous interpretation of modulo-two pi phase measurements. Curtis Chen has recently developed a method based on maximum likelihood statistics and network flow models -- see his thesis and a sample figure.
- Atmospheric propagation and correction
When radar signals travel through the atmosphere, they are delayed by the presence of water vapor. This corrupts the phase measurements in interferograms. Fayaz Onn is studying this phenomenon and how InSAR data may be improved through combination with GPS data. See a corrupted interferogram.
- High speed computing and networking
Radar systems generate very large sets of data, and processing these in a timely way is important not only for data throughput but also because development of advanced radar often requires many iterations to produce products with the accuracies and capabilities we seek. We implement our codes on high speed and parallel machines, often beginning with prototyping in a package such as Matlab that converts easily to Fortran 90 and subsequently to parallelized codes. Efficient code generation underlies much of our work.