Dealing with artefacts In JET iterative bolometric tomography using masks

Bolometric tomography is a widely applied technique to infer important indirect quantities in magnetically confined plasmas, such as the total radiated power. However, being an inverse and ill-posed problem, the tomographic algorithms have to be carefully steered to converge on the most approriate solutions and often specialists have to balance the quality of the obtained reconstructions between the core and the edge of the plasma. Given the topology of the emission and the layout of the diagnostics in practically all devices, the tomographic inversions of bolometry are often affected by artefacts, which can influence derived quantities and specific studies based on the reproduced tomograms, such as power balance studies and benchmarching of gyrokinetic simulations. This article deals with the introduction of a simple, but very efficient methodology. It is based on constraining the solution of the tomographic inversions by using a specific estimate of the initial solution, built with the data from specific combinations of detectors (called ‘masks’). It has been tested with phantoms and with real data, using the Maximum Likelihood approach at JET. Results show how the obtained tomograms improve sensibly both in the core and at the edge of the device when compared with those obtained without the use of masks as initial guess. The correction for the main artefacts can have a significant impact on the interpretation of both the core (electron transport, alpha heating) and the edge physics (detachment , SOL). The method is completely general and can be applied by any iterative algorithm starting from an initial guess for the emission profile to be reconstructed.

Direct P-Wave Travel Time Tomography of the Middle East

<p>High-resolution seismic images of the crust and mantle beneath regions of complex surface geological structures are necessary to gain insights on the underlying geodynamical processes. One such region embodying various plate boundary motions and intraplate deformations is the Middle East, and consequently the region is prone to significant seismic activity. Hence a tomographic investigation using a more recent and reliable data set is vital in understanding the ongoing complicated deformation process driven by the African, Arabian and Eurasian plates. The purpose of our study is to retrieve a detailed  model of the crust and mantle beneath the Middle Eastern region using teleseismic P arrival times from the ISC-EHB bulletin (Engdahl et al., 1998).</p><p>Starting with AK135 as the reference model we invert for tomographic models of compressional wavespeed perturbations down to lower mantle depths in an area bounded by longitudes 22E–66E and latitudes 8N–48N.  The data set used in this study consists of regionally observed P-phase arrival times from over 1000 global events from 1996–2016 culminating in a larger dataset than other similar studies. Selection of a reliable data, ray tracing, preconditioning and inversion steps are carried out using the BD-soft software suite (https://www.geoazur.fr/GLOBALSEIS/Soft.html).</p><p>Preliminary inversion results are consistent with the previous regional tomographic studies. In checkerboard tests, cell sizes as low as ∼ 2.8° × 2.8° ( ∼ 240 × 240 km at surface) are generally well recovered down to a 1000 km depth beneath the Anatolian plateau where we currently have the densest coverage. Additionally the Caucasus region and northern parts of the Iranian plateau shows good recovery of ±4% Vp perturbation amplitudes at depths ∼ 70 – 135 km. There is fair recovery for a minimum cell size of ∼ 2.8° × 2.8° beneath the Iranian Plateau, Zagros mountain region, Persian gulf, and northeast Iraq, along with quite good recovery of cell amplitudes towards the Anatolian-Caucasus region at depth ranges 380 – 430 km, 650 – 700 km, and around 950 km. Tomographic inversions unveil a low P velocity zone stretching from the Afar region to Sinai Peninsula consistent with S wave velocity observations of a similar feature by Chang and van der Lee 2011.</p><p>We are able to further improve coverage especially down to lithospheric depths within the Arabian peninsula using first arrival times measured from waveform data collected from regional networks. Addition of first arrival time delays from waveforms highlights a prominent low velocity in the tomographic inversions beneath the volcanic fields of western Saudi Arabia. Our ultimate goal is to perform full-waveform inversion of the region constrained by the constructed P-wave model.</p>

An Adjoint Technique for Estimation of Interstation Phase and Group Dispersion from Ambient Noise Cross Correlations

A method of extracting group and phase velocity dispersions jointly for Love‐ and Rayleigh‐wave observations is presented. This method uses a spectral element representation of a path average Earth model parameterized with density, shear‐wave velocity, radial anisotropy, and VP/VS ratio. An initial dispersion curve is automatically estimated using a heuristic approach to prevent misidentification of the phase. A second step then more accurately fits the observed noise correlation function (NCF) between interstation pairs in the frequency domain. For good quality cross correlations with reasonable signal‐to‐noise ratio, we are able to very accurately fit the spectrum of NCFs and hence obtain reliable estimates of both phase and group velocity jointly for Love and Rayleigh surface waves. In addition, we also show how uncertainties can be estimated with linearized approximations from the Jacobians and subsequently used in tomographic inversions.

Optimization of soft X-ray tomography on the COMPASS tokamak

The COMPASS tokamak is equipped with the soft X-ray (SXR) diagnostic system based on silicon photodiode arrays shielded by a thin beryllium foil. The diagnostic is composed of two pinhole cameras having 35 channels each and one vertical pinhole camera with 20 channels, which was installed recently to improve tomographic inversions. Lines of sight of the SXR detectors cover almost complete poloidal cross section of the COMPASS vessel with a spatial resolution of 1-2 cm and temporal resolution of about 3 μs. Local emissivity is reconstructed via Tikhonov regularization constrained by minimum Fisher information that provides reliable and robust solution despite limited number of projections and ill-conditionality of this task. Improved border conditions and numerical differentiation matrices suppressing artifacts in reconstructed radiation were implemented in the code. Furthermore, a fast algorithm eliminating iterative processes was developed, and it is foreseen to be tested in real-time plasma control.

Ionospheric precursors to scintillation activity

<p>Ionospheric scintillation is the rapid fluctuation of both phase and amplitude of trans-ionospheric radio waves due to small scale electron density irregularities in the ionosphere. Prediction of the occurrence of scintillation at L band frequencies is needed to mitigate the disruption of space-based communication and navigation systems. The purpose of this paper is to present a method of using tomographic inversions of the ionospheric electron density obtained from ground-based GPS data to infer the location and strength of the post-sunset plasma drift vortex. This vortex is related to the pre-reversal enhancement in the eastwards electric field which has been correlated to the subsequent occurrence of scintillation.</p>

Study on the limitations of travel-time inversion applied to sub-basalt imaging

The difficulties of seismic imaging beneath high velocity structures are widely recognised. In this setting, theoretical analysis of synthetic wide-angle seismic reflection data indicates that velocity models are not well constrained. A two-dimensional velocity model was built to simulate a simplified structural geometry given by a basaltic wedge placed within a sedimentary sequence. This model reproduces the geological setting in areas of special interest for the oil industry as the Faroe-Shetland Basin. A wide-angle synthetic dataset was calculated on this model using an elastic finite difference scheme. This dataset provided travel times for tomographic inversions. Results show that the original model can not be completely resolved without considering additional information. The resolution of nonlinear inversions lacks a functional mathematical relationship, therefore, statistical approaches are required. Stochastic tests based on Metropolis techniques support the need of additional information to properly resolve sub-basalt structures.

Study on the limitations of traveltime inversion in the presence of extreme velocity anomalies

The difficulties of seismic imaging beneath high velocity structures are widely recognised. In this setting, theoretical analysis of synthetic wide-angle seismic reflection data indicates that velocity models are not well constrained. A two-dimensional velocity model was built to simulate a simplified structural geometry given by a basaltic wedge placed within a sedimentary sequence. This model reproduces the geological setting in areas of special interest for the oil industry as the Faroe-Shetland Basin. A wide-angle synthetic dataset was calculated on this model using an elastic finite difference scheme. This dataset provided travel times for tomographic inversions. Results show that the original model can not be completely resolved without considering additional information. The resolution of nonlinear inversions lacks a functional mathematical relationship, therefore, statistical approaches are required. Stochastical tests based on Metropolis techniques support the need of additional information to properly resolve subbasalt structures.

Estimation of electron densities in the lower thermosphere from GUVI 135.6 nm tomographic inversions in support of SpreadFEx

The SpreadFEx campaign was conducted with the goal of investigating potential neutral atmospheric dynamics influences in seeding plasma instabilities and bubbles extending to higher altitudes from September to November 2005, with primary measurements in Brazil. In this paper, we present the results of space-based UV and ground-based optical observations in support of this campaign. Specifically, we present multi-dimensional electron density images obtained tomographically from the 135.6 nm emissions measured by the GUVI instrument aboard the TIMED satellite that result from radiative recombination of O+ and compare those with the corresponding 630.0 nm OI images recorded in the Brazilian sector. The GUVI results provide altitude vs. longitude information on depleted regions in the ionospheric plasma density that are complementary to the single-height latitude-longitude images obtained with the airglow imager.

IDENTIFYING MODAL ARRIVALS IN SHALLOW WATER FOR BOTTOM GEOACOUSTIC INVERSIONS

A mode identification process that can be applied to broadband acoustic transmissions in the ocean is presented. The process is associated with a modal travel time inversion scheme for geoacoustic or tomographic inversions. The process is based on the assumption that a reference (background) environment is known and the identification process is based on information on the group (modal) velocities of the reference environment only. Using two characteristic examples corresponding to shallow water environments it is shown that the identification process works well at least for the lower order modes.

Topography can affect linearization in tomographic inversions

Linearized inverse techniques commonly are used to solve for velocity models from traveltime data. The amount that a model may change without producing large, nonlinear changes in the predicted traveltime data is dependent on the surface topography and parameterization. Simple, one‐layer, laterally homogeneous, constant‐gradient models are used to study analytically and empirically the effect of topography and parameterization on the linearity of the model‐data relationship. If, in a weak‐velocity‐gradient model, rays turn beneath a valley with topography similar to the radius of curvature of the raypaths, then large nonlinearities will result from small model perturbations. Hills, conversely, create environments in which the data are more nearly linearly related to models with the same model perturbations.