Data filtering for thermal mapping of road surface temperatures

Abstract

Abstract‘Thermal mapping’ is the name given to the measurement of the variation in road surface temperatures across a region on winter nights. The method involves the use of infrared thermometers and data loggers mounted in modified vehicles. The signal recorded by the data logger is composed of two parts: that due to the road surface temperature and that due to error. Because relative temperature differences, as opposed to absolute temperatures, are important for thermal mapping, a low pass filter to remove or reduce high‐frequency errors or ‘noise’ is required. This paper examines the required features of such a filter and tests the applicability of one such filter with a thermal mapping data set. The results show that high‐frequency noise may be reduced without losing the spatial resolution necessary to detect cold spots in the road network. It is also demonstrated that in order to achieve this goal, a sufficiently high spatial resolution of raw observations is required.