Mapping of thermal fields in fluids by interferometry has been describes in the present article. By interpreting the interferograms as path integrals of the field variable, it has been shown that the three-dimensional field can be reconstructed using principles of tomography. Iterative algorithms based on the ART family have been sound to be suitable for this purpose. Experimental results for buoyancy-driven convection in three configurations namely, a protruding heater, a two dimensional square cavity, and a differently heated horizontal fluid layer have been presented.

Interferometric tomography has good potential for applications to problems of far greater complexity. Examples are: (1) fully unstady three-dimensional flow, (2) simultaneous reconstruction of fluid-fluid interfaces along with the temperature variation over them, (3) strongly refracting thermal fields where the reconstruction has to be performed with integrals evaluated over space curves, and (4) monitoring the growth of laser crystals from an aqueous solution by controlling the prevailing thermal fields. There is also the possibility of recovering the velocity field in the fluid medium from the complete temperature data. With advances being made in improving the spatial resolution and recording speed beyond video rates, it is now conceivable that interferometric tomography can be used to analyze turbulance structures, for example in buoyant flows, jets, and chemically reacting fluids. Color interferometry is a promising tool to improve spatial resolution since fringes will form corresponding to each wavelength of the laser. Tomography by itself is a versatile tool to analyze data recorded by other measurement strategies that do not rely on changes in the refractive index. Projection data for example, can also be generated by reflaction, absorption, and attenuation mechanism.

Of great practical utility would be the extension of laser tomography to field-scale problems. Optical elements are needed here mainly to generate the projection data. Thus a field-scale problem is characterized by a large volume of data and the processing reduces to purely a numerical challenge. This aspect will be resolved with further improvements in computer technology. The use of lasers to process satellite images, and in turn provide a reference input to weather prediction codes, is an application that can deliver significant benifits to the society.