Source identification from line integral measurements and simple atmospheric models (Q1951326)
From MaRDI portal
| This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use this page instead for the normal view: Source identification from line integral measurements and simple atmospheric models |
scientific article; zbMATH DE number 6170740
| Language | Label | Description | Also known as |
|---|---|---|---|
| English | Source identification from line integral measurements and simple atmospheric models |
scientific article; zbMATH DE number 6170740 |
Statements
Source identification from line integral measurements and simple atmospheric models (English)
0 references
5 June 2013
0 references
Mathematically speaking, this paper concerns the problem of reconstruction of the initial condition \[ u(x,0) = u_0(x) \] in some domain \(\Omega\), where \(u(x,t)\) satisfies the advection-diffusion equation \[ u_t = \nabla\cdot (a(x) \nabla u) - \psi(x)\cdot \nabla u \] with Dirichlet or Neumann boundary conditions. Here \(a(x)\) and \(\psi(x)\) are assumed to be given. The data used to reconstruct the initial condition are the line integrals of the solution at a given time \(t = T\) along some lines \(l_i\), \(i = 1, \dots, m\): \[ b_i = \int_{l_i} u(x,T) ds. \] The authors consider the cases when the intial condition are given as single or mutliple delta functions, i.e., the case when the solution \(u\) is generated by point sources. A least squares approach with adjoint problem is used to solve the problem. Several numerical examples with simulated data are presented.
0 references
advection-diffusion equation
0 references
compressed sensing
0 references
computed tomography
0 references
sparsity
0 references
standoff chemical detection
0 references