Ranjan, Shashi
(2016)
3-D Electrical Resistivity Imaging with Compressive Sensing: Application to Simulate Soil-Water-Disease Interactions of Citrus Trees.
Masters thesis, Indian Institute of Technology Hyderabad.
Abstract
This research is organized into two parts. First part of the work aims at understanding
the inversion mechanism of electrical resistivity tomography (ERT) from first principles.
A compressive sensing (CS) based ERT inversion algorithm was developed
for use with sub-surface image reconstruction. Second part of the research aims at
developing efficient irrigation water management scenarios considering health of the
citrus tree using experimental and numerical studies.
Geophysical techniques are widely used to characterize hydrological
uxes that
are controlling sub-surface dynamics. Of these, ERT has proven to be the prominent
and robust technologies for imaging the sub-surface. The strength of a geophysical
inversion technique to reconstruct the image is largely dependent on the strategy
adopted to solve the ill-posed, under-determined, non-linear system. Conventional
gradient based algorithms that use Euclidean norm minimization with some sort of
regularization may fail to detect the sharp interfaces between earth layers and resistivity
anomalies. Given the fact that, earth's resistivity (and hence, hydro-geologic)
features varies abruptly on a continuous spatial domain; the sparsity in model parameter
change can be better utilized to apply CS algorithms for resistivity imaging. CS
based 3-D image reconstruction algorithms from ERT observations were not reported
in literature till date.
Actions (login required)
 |
View Item |
Altmetric
Altmetric