New directions in sparse sampling and estimation for underdetermined systems

A central objective in signal processing is to infer meaningful information from a set of measurements or data. While most signal models have an overdetermined structure (the number of unknowns less than the number of equations), traditionally very few statistical estimation problems have considered a data model which is underdetermined (number of unknowns more than the number of equations). However, in recent times, an explosion of theoretical and computational methods have been developed primarily to study underdetermined systems by imposing sparsity on the unknown variables. This is motivated by the observation that inspite of the huge volume of data that arises in sensor networks, genomics, imaging, particle physics, web search etc., their information content is often much smaller compared to the number of raw measurements. This has given rise to the possibility of reducing the number of measurements by down sampling the data, which automatically gives rise to underdetermined systems.

In this thesis, we provide new directions for estimation in an underdetermined system, both for a class of parameter estimation problems and also for the problem of sparse recovery in compressive sensing. There are two main contributions of the thesis: design of new sampling and statistical estimation algorithms for array processing, and development of improved guarantees for sparse reconstruction by introducing a statistical framework to the recovery problem.

We consider underdetermined observation models in array processing where the number of unknown sources simultaneously received by the array can be considerably larger than the number of physical sensors. We study new sparse spatial sampling schemes (array geometries) as well as propose new recovery algorithms that can exploit priors on the unknown signals and unambiguously identify all the sources. The proposed sampling structure is generic enough to be extended to multiple dimensions as well as to exploit different kinds of priors in the model such as correlation, higher order moments, etc.

Recognizing the role of correlation priors and suitable sampling schemes for underdetermined estimation in array processing, we introduce a correlation aware framework for recovering sparse support in compressive sensing. We show that it is possible to strictly increase the size of the recoverable sparse support using this framework provided the measurement matrix is suitably designed. The proposed nested and coprime arrays are shown to be appropriate candidates in this regard. We also provide new guarantees for convex and greedy formulations of the support recovery problem and demonstrate that it is possible to strictly improve upon existing guarantees.

This new paradigm of underdetermined estimation that explicitly establishes the fundamental interplay between sampling, statistical priors and the underlying sparsity, leads to exciting future research directions in a variety of application areas, and also gives rise to new questions that can lead to stand-alone theoretical results in their own right.

I. Impact spallation experiments: fracture patterns and spall velocities. II. Craters in carbonate rocks: an electron paramagnetic resonance analysis of shock damage

This work is divided into two independent papers.

PAPER 1.

Spall velocities were measured for nine experimental impacts into San Marcos gabbro targets. Impact velocities ranged from 1 to 6.5 km/sec. Projectiles were iron, aluminum, lead, and basalt of varying sizes. The projectile masses ranged from a 4 g lead bullet to a 0.04 g aluminum sphere. The velocities of fragments were measured from high-speed films taken of the events. The maximum spall velocity observed was 30 m/sec, or 0.56 percent of the 5.4 km/sec impact velocity. The measured velocities were compared to the spall velocities predicted by the spallation model of Melosh (1984). The compatibility between the spallation model for large planetary impacts and the results of these small scale experiments are considered in detail.

The targets were also bisected to observe the pattern of internal fractures. A series of fractures were observed, whose location coincided with the boundary between rock subjected to the peak shock compression and a theoretical "near surface zone" predicted by the spallation model. Thus, between this boundary and the free surface, the target material should receive reduced levels of compressive stress as compared to the more highly shocked region below.

PAPER 2.

Carbonate samples from the nuclear explosion crater, OAK, and a terrestrial impact crater, Meteor Crater, were analyzed for shock damage using electron para- magnetic resonance, EPR. The first series of samples for OAK Crater were obtained from six boreholes within the crater, and the second series were ejecta samples recovered from the crater floor. The degree of shock damage in the carbonate material was assessed by comparing the sample spectra to spectra of Solenhofen limestone, which had been shocked to known pressures.

The results of the OAK borehole analysis have identified a thin zone of highly shocked carbonate material underneath the crater floor. This zone has a maximum depth of approximately 200 ft below sea floor at the ground zero borehole and decreases in depth towards the crater rim. A layer of highly shocked material is also found on the surface in the vicinity of the reference bolehole, located outside the crater. This material could represent a fallout layer. The ejecta samples have experienced a range of shock pressures.

It was also demonstrated that the EPR technique is feasible for the study of terrestrial impact craters formed in carbonate bedrock. The results for the Meteor Crater analysis suggest a slight degree of shock damage present in the β member of the Kaibab Formation exposed in the crater walls.

Experimental chemical physics of droplets and clusters

Time-of-flight measurements of energetic He atoms, field ionization of cryogenic liquid helium clusters, and time-of-flight and REMPI spectroscopy of radical salt clusters were investigated experimentally. The excited He atoms were generated in a corona discharge. Two strong neutral peaks were observed, accompanied by a prompt photon peak and a charged peak. All peaks were correlated with the pulsing of the discharge. The neutral hyperthermal and metastable atoms were formed by different mechanisms at different stages of the corona discharge. Positively charged helium droplets were produced by ionization of liquid helium in an electrostatic spraying experiment. The fluid emerging from a thin glass capillary was ionized by a high voltage applied to a needle inside the capillary. Fine droplets (less than 10 µm in diameter) were produced in showers with currents as high as 0.4 µA at 2-4 kV. The high currents resulting from field ionization in helium and the low surface tension of He I, led to charge densities that greatly exceeded the Rayleigh limit, thus resulting in coulombic explosion of the liquid. In contrast, liquid nitrogen formed a well-defined Taylor cone with droplets having diameters comparable to the jet (≈100 µm) at lower currents (10 nA) and higher voltages (8 kV). The metal-halide clusters of calcium and chlorine were generated by laser ablation of calcium metal in a Ar/CCl₄ expansion. A visible spectrum of the Ca₂Cl₃ cluster was observed from 651 to 630 nm by 1 +1' REMPI. The spectra were composed of a strong origin band at 15 350.8 cm^-1 and several weak vibronic bands. Density functional calculations predicted three minimum energy isomers. The spectrum was assigned to the ²B₂ ← X ²A₁ transition of a planar C_2V structure having a ring of two Cl and two Ca atoms and a terminal Cl atom. The ring isomer of Ca₂Cl₃ has the unpaired electron localized on one Ca²⁺ ion to form a Ca⁺ chromophore. A second electronic band of Ca₂Cl₃ was observed at 720 nm. The band is sharply different from the 650 nm band and likely due to a different isomer.

I. The Superstition Hills, California, earthquakes of 24 November 1987. II. Three-dimensional velocity structure of southern California.

Part 1 of this thesis is about the 24 November, 1987, Superstition Hills earthquakes. The Superstition Hills earthquakes occurred in the western Imperial Valley in southern California. The earthquakes took place on a conjugate fault system consisting of the northwest-striking right-lateral Superstition Hills fault and a previously unknown Elmore Ranch fault, a northeast-striking left-lateral structure defined by surface rupture and a lineation of hypocenters. The earthquake sequence consisted of foreshocks, the M_s 6.2 first main shock, and aftershocks on the Elmore Ranch fault followed by the M_s 6.6 second main shock and aftershocks on the Superstition Hills fault. There was dramatic surface rupture along the Superstition Hills fault in three segments: the northern segment, the southern segment, and the Wienert fault.

In Chapter 2, M_L≥4.0 earthquakes from 1945 to 1971 that have Caltech catalog locations near the 1987 sequence are relocated. It is found that none of the relocated earthquakes occur on the southern segment of the Superstition Hills fault and many occur at the intersection of the Superstition Hills and Elmore Ranch faults. Also, some other northeast-striking faults may have been active during that time.

Chapter 3 discusses the Superstition Hills earthquake sequence using data from the Caltech-U.S.G.S. southern California seismic array. The earthquakes are relocated and their distribution correlated to the type and arrangement of the basement rocks. The larger earthquakes occur only where continental crystalline basement rocks are present. The northern segment of the Superstition Hills fault has more aftershocks than the southern segment.

An inversion of long period teleseismic data of the second mainshock of the 1987 sequence, along the Superstition Hills fault, is done in Chapter 4. Most of the long period seismic energy seen teleseismically is radiated from the southern segment of the Superstition Hills fault. The fault dip is near vertical along the northern segment of the fault and steeply southwest dipping along the southern segment of the fault.

Chapter 5 is a field study of slip and afterslip measurements made along the Superstition Hills fault following the second mainshock. Slip and afterslip measurements were started only two hours after the earthquake. In some locations, afterslip more than doubled the coseismic slip. The northern and southern segments of the Superstition Hills fault differ in the proportion of coseismic and postseismic slip to the total slip.

The northern segment of the Superstition Hills fault had more aftershocks, more historic earthquakes, released less teleseismic energy, and had a smaller proportion of afterslip to total slip than the southern segment. The boundary between the two segments lies at a step in the basement that separates a deeper metasedimentary basement to the south from a shallower crystalline basement to the north.

Part 2 of the thesis deals with the three-dimensional velocity structure of southern California. In Chapter 7, an a priori three-dimensional crustal velocity model is constructed by partitioning southern California into geologic provinces, with each province having a consistent one-dimensional velocity structure. The one-dimensional velocity structures of each region were then assembled into a three-dimensional model. The three-dimension model was calibrated by forward modeling of explosion travel times.

In Chapter 8, the three-dimensional velocity model is used to locate earthquakes. For about 1000 earthquakes relocated in the Los Angeles basin, the three-dimensional model has a variance of the the travel time residuals 47 per cent less than the catalog locations found using a standard one-dimensional velocity model. Other than the 1987 Whittier earthquake sequence, little correspondence is seen between these earthquake locations and elements of a recent structural cross section of the Los Angeles basin. The Whittier sequence involved rupture of a north dipping thrust fault bounded on at least one side by a strike-slip fault. The 1988 Pasadena earthquake was deep left-lateral event on the Raymond fault. The 1989 Montebello earthquake was a thrust event on a structure similar to that on which the Whittier earthquake occurred. The 1989 Malibu earthquake was a thrust or oblique slip event adjacent to the 1979 Malibu earthquake.

At least two of the largest recent thrust earthquakes (San Fernando and Whittier) in the Los Angeles basin have had the extent of their thrust plane ruptures limited by strike-slip faults. This suggests that the buried thrust faults underlying the Los Angeles basin are segmented by strike-slip faults.

Earthquake and explosion travel times are inverted for the three-dimensional velocity structure of southern California in Chapter 9. The inversion reduced the variance of the travel time residuals by 47 per cent compared to the starting model, a reparameterized version of the forward model of Chapter 7. The Los Angeles basin is well resolved, with seismically slow sediments atop a crust of granitic velocities. Moho depth is between 26 and 32 km.

Part I - The radiation of elastic waves from a spherical cavity in a half space. Part II - Precision determination of focal depths and epicenters of earthquakes.

Part I: The dynamic response of an elastic half space to an explosion in a buried spherical cavity is investigated by two methods. The first is implicit, and the final expressions for the displacements at the free surface are given as a series of spherical wave functions whose coefficients are solutions of an infinite set of linear equations. The second method is based on Schwarz's technique to solve boundary value problems, and leads to an iterative solution, starting with the known expression for the point source in a half space as first term. The iterative series is transformed into a system of two integral equations, and into an equivalent set of linear equations. In this way, a dual interpretation of the physical phenomena is achieved. The systems are treated numerically and the Rayleigh wave part of the displacements is given in the frequency domain. Several comparisons with simpler cases are analyzed to show the effect of the cavity radius-depth ratio on the spectra of the displacements.

Part II: A high speed, large capacity, hypocenter location program has been written for an IBM 7094 computer. Important modifications to the standard method of least squares have been incorporated in it. Among them are a new way to obtain the depth of shocks from the normal equations, and the computation of variable travel times for the local shocks in order to account automatically for crustal variations. The multiregional travel times, largely based upon the investigations of the United States Geological Survey, are confronted with actual traverses to test their validity.

It is shown that several crustal phases provide control enough to obtain good solutions in depth for nuclear explosions, though not all the recording stations are in the region where crustal corrections are considered. The use of the European travel times, to locate the French nuclear explosion of May 1962 in the Sahara, proved to be more adequate than previous work.

A simpler program, with manual crustal corrections, is used to process the Kern County series of aftershocks, and a clearer picture of tectonic mechanism of the White Wolf fault is obtained.

Shocks in the California region are processed automatically and statistical frequency-depth and energy depth curves are discussed in relation to the tectonics of the area.

Supernovae explosions induced by pair production instability

Stars with a core mass greater than about 30 M_⊙ become dynamically unstable due to electron-positron pair production when their central temperature reaches 1.5-2.0 x 10^{9 0}K. The collapse and subsequent explosion of stars with core masses of 45, 52, and 60 M_⊙ is calculated. The range of the final velocity of expansion (3,400 – 8,500 km/sec) and of the mass ejected (1 – 40 M_⊙) is comparable to that observed for type II supernovae.

An implicit scheme of hydrodynamic difference equations (stable for large time steps) used for the calculation of the evolution is described.

For fast evolution the turbulence caused by convective instability does not produce the zero entropy gradient and perfect mixing found for slower evolution. A dynamical model of the convection is derived from the equations of motion and then incorporated into the difference equations.

Cosmic Explosions: Observations Of Infant Hydrogen-Free Supernovae Towards An Understanding Of Their Parent Systems

Radiation in the first days of supernova explosions contains rich information about physical properties of the exploding stars. In the past three years, I used the intermediate Palomar Transient Factory to conduct one-day cadence surveys, in order to systematically search for infant supernovae. I show that the one-day cadences in these surveys were strictly controlled, that the realtime image subtraction pipeline managed to deliver transient candidates within ten minutes of images being taken, and that we were able to undertake follow-up observations with a variety of telescopes within hours of transients being discovered. So far iPTF has discovered over a hundred supernovae within a few days of explosions, forty-nine of which were spectroscopically classified within twenty-four hours of discovery.

Our observations of infant Type Ia supernovae provide evidence for both the single-degenerate and double-degenerate progenitor channels. On the one hand, a low-velocity Type Ia supernova iPTF14atg revealed a strong ultraviolet pulse within four days of its explosion. I show that the pulse is consistent with the expected emission produced by collision between the supernova ejecta and a companion star, providing direct evidence for the single degenerate channel. By comparing the distinct early-phase light curves of iPTF14atg to an otherwise similar event iPTF14dpk, I show that the viewing angle dependence of the supernova-companion collision signature is probably responsible to the difference of the early light curves. I also show evidence for a dark period between the supernova explosion and the first light of the radioactively-powered light curve. On the other hand, a peculiar Type Ia supernova iPTF13asv revealed strong near-UV emission and absence of iron in the spectra within the first two weeks of explosion, suggesting a stratified ejecta structure with iron group elements confined to the slow-moving part of the ejecta. With its total ejecta mass estimated to exceed the Chandrasekhar limit, I show that the stratification and large mass of the ejecta favor the double-degenerate channel.

In a separate approach, iPTF found the first progenitor system of a Type Ib supernova iPTF13bvn in the pre-explosion HST archival mages. Independently, I used the early-phase optical observations of this supernova to constrain its progenitor radius to be no larger than several solar radii. I also used its early radio detections to derive a mass loss rate of 3e-5 solar mass per year for the progenitor right before the supernova explosion. These constraints on the physical properties of the iPTF13bvn progenitor provide a comprehensive data set to test Type Ib supernova theories. A recent HST revisit to the iPTF13bvn site two years after the supernova explosion has confirmed the progenitor system.

Moving forward, the next frontier in this area is to extend these single-object analyses to a large sample of infant supernovae. The upcoming Zwicky Transient Facility with its fast survey speed, which is expected to find one infant supernova every night, is well positioned to carry out this task.