The nature of continuous fields implies the necessity to study them through sampling. Further, spatial interpolation allows the estimation of values for points that were originally discarded. Purely spatial interpolation disregards the temporal component of captured data. Such an approach, however, may lead to wrong conclusions if the data being studied is change-prone. This paper analyses spatio-temporal interpolation of air temperature data captured by an unmanned aerial vehicle (UAV). Since each point in the resulting dataset has a unique timestamp, purely spatial interpolation of UAV sensor data is likely to be less accurate than spatio-temporal interpolation. In the case of temperature data, for instance, spatio-temporal interpolation makes it possible to monitor the movement of pockets of warm air. Geoinformation systems, like ArcGIS or QGIS, have rich spatial interpolation toolsets. These systems, however, provide no native spatio-temporal interpolation means. We explored spatio-temporal interpolation of UAV data and developed a prototype of a stand-alone interpolation tool that exploits radial basis functions (RBF) and inverse distance weighting (IDW) for interpolation in a continuous space-time domain. This paper discusses spatio-temporal interpolation of UAV data in comparison with purely spatial interpolation, as well as the application of the developed tool.