Quantitative reconstruction of thermal and dynamic characteristics of lava flow from surface thermal measurements. Issue 3 (3rd April 2016)

• Record Type: Journal Article
• Authors: Korotkii, Alexander; Kovtunov, Dmitry; Ismail-Zadeh, Alik; Tsepelev, Igor; Melnik, Oleg
• Is Part Of: Geophysical journal international. Volume 205:Issue 3(2016:Jun.)
• Page Start: 1767
• Page End: 1779
• Publication Date: 2016-04-03
• Subjects: Numerical solutions -- Inverse theory -- Numerical approximations and analysis -- Effusive volcanism; Geophysics -- Periodicals; 550
• DOI: 10.1093/gji/ggw117
• Languages: English
• ISSNs: 0956-540X

Abstract

Abstract: We study a model of lava flow to determine its thermal and dynamic characteristics from thermal measurements of the lava at its surface. Mathematically this problem is reduced to solving an inverse boundary problem. Namely, using known conditions at one part of the model boundary we determine the missing condition at the remaining part of the boundary. We develop a numerical approach to the mathematical problem in the case of steady-state flow. Assuming that the temperature and the heat flow are prescribed at the upper surface of the model domain, we determine the flow characteristics in the entire model domain using a variational (adjoint) method. We have performed computations of model examples and showed that in the case of smooth input data the lava temperature and the flow velocity can be reconstructed with a high accuracy. As expected, a noise imposed on the smooth input data results in a less accurate solution, but still acceptable below some noise level. Also we analyse the influence of optimization methods on the solution convergence rate. The proposed method for reconstruction of physical parameters of lava flows can also be applied to other problems in geophysical fluid flows.