Measuring heat flux on the surface of a natural reservoir by its infrared image, taking into account environmental conditions

When convection is thermal, there is no insolation, and water is fresh, the upward heat flux on the surface of a natural reservoir can be presented as the product of the root-mean-squire deviation (RMSD) of water-surface temperature from its mean in the power of 4/3 and a polynomial function of the Prandtl number, where the latter corresponds to the bulk water temperature. The error of measuring the heat flux arising due to this representation is estimated. To be applicable in the sea, the proposed heat-flux representation is corrected, so, that to take into account water salinity, which becomes essential when thermal convection develops in cold water. In the Prandtl number range 4.4-7.2, the skewness of the surface temperature, which, along with its RMSD from the mean, can be measured by infrared imaging, is a linear function of the Prandtl number (Verevochkin [4], Int J Heat Mass Transfer 143:118426). It is shown that, in this case, the relative error of calculating the Prandtl number is less than occurring while measuring the skewness. A possibility of using the spatial-average water-surface temperature to calculate the Prandtl number is also considered, and the relative error arising here is estimated. Two heat-flux values corresponding to the same conditions, but obtained in different ways, are used as coordinates of a point. The first value measured calorimetrically is taken from literature sources, while the second one is calculated by the being proposed here formulae. The linear function fitted to the set of thirteen such points by the least square method has the slope 1.15 +/- 0.11.