Hydrology and Land Surface Studies
Moment inference from tomograms
Article first published online: 22 NOV 2007
Copyright 2007 by the American Geophysical Union.
Geophysical Research Letters
Volume 34, Issue 22, November 2007
How to Cite
2007), Moment inference from tomograms, Geophys. Res. Lett., 34, L22404, doi:10.1029/2007GL031621., , and (
- Issue published online: 22 NOV 2007
- Article first published online: 22 NOV 2007
- Manuscript Accepted: 22 OCT 2007
- Manuscript Revised: 10 OCT 2007
- Manuscript Received: 17 AUG 2007
 Time-lapse geophysical tomography can provide valuable qualitative insights into hydrologic transport phenomena associated with aquifer dynamics, tracer experiments, and engineered remediation. Increasingly, tomograms are used to infer the spatial and/or temporal moments of solute plumes; these moments provide quantitative information about transport processes (e.g., advection, dispersion, and rate-limited mass transfer) and controlling parameters (e.g., permeability, dispersivity, and rate coefficients). The reliability of moments calculated from tomograms is, however, poorly understood because classic approaches to image appraisal (e.g., the model resolution matrix) are not directly applicable to moment inference. Here, we present a semi-analytical approach to construct a moment resolution matrix based on (1) the classic model resolution matrix and (2) image reconstruction from orthogonal moments. Numerical results for radar and electrical-resistivity imaging of solute plumes demonstrate that moment values calculated from tomograms depend strongly on plume location within the tomogram, survey geometry, regularization criteria, and measurement error.