**5. Discussion**

Based on the research and analysis, it was found that the value of the dielectric constant determined continuously based on the amplitude of the wave reflected from the surface is a ffected by the atmospheric conditions on the surface—its temperature, moisture, ice or snow, and presence of de-icing salt.

When the temperature is high, HMA dielectric constants values are higher than at temperatures below zero. The observed tendency of the apparent increase in HMA dielectric constants with the increase in its temperature results from the more energetic movement of molecules—this enables an easier formation and orientation of dipoles and requires continuing testing on a larger number of samples and/or test sites in a continuous temperature range. As a result of the presence of a water film on the surface and partly in the pores of the HMA (water itself is a dipole), after the application of an external electric field, as expected, an increase in the polarization capacity of the medium was observed. The freezing of water on the surface and partly in the pores caused a slight decrease in the dielectric constant. This is due to the formation of hydrogen bonds and retention of the dipole polarization mechanism of water. An increase in dielectric constant, as it was observed, was due to the presence of snow on the surface. The apparent increase in HMA dielectric constant due to the presence of salt can be combined with the presence of Na− and Cl+ ions as well as ionic polarization and increase in the material's ability to compensate the electric field; on the other hand, the electric field between the Na− and Cl+ ions and water molecules whose presence results from melting snow and ice on the surface of the slab should level out. This issue as well as the issue related to the presence of snow on the surface should be further investigated.

The measurement error in calculating the thickness based on reflection amplitude in the conditions adopted as a reference is up to −10%. The surface atmospheric conditions a ffect less or more the thickness value determined by the GPR method—and as a result of the thickness determination based on the dielectric constant calculated using the amplitude of the reflected wave from the surface below zero temperature, the thickness measurement error was up to −11 ÷ +8%; in the case of a wave amplitude reflected from the surface covered with water film, it was up to −11 ÷ +7%; in the case of a wave amplitude reflected from icy surfaces, it was up to −7 ÷ 8%; in the case of a wave amplitude reflected from the surfaces covered with a thin layer of powdery snow, it was up to −16 ÷ +10%, and in the case of a wave amplitude reflected from the surfaces covered with de-icing salt, it was up to −29% (see Table 9).

It was proposed to use advanced wavelet analysis to determine which atmospheric conditions on the surface of the pavement result in high/low dielectric constant values determined based on the amplitude. Promisingly, high di fferences in magnitudes of scale factors in relation to the reference conditions were obtained based on a signal from the measurements on the surface covered with de-icing salt, which is a very positive result, because the presence of salt on the surface is not observable on video records performed alongside GPR measurements.

For each of the analyzed atmospheric conditions, a correction factor for the dielectric constant calculated based on the wave amplitudes reflected from surfaces with di fferent atmospheric conditions was proposed, whose value could be chosen based on wavelet analysis and exactly reflects the appearance of the wavelet scalogram (see Figures 8–10). During determinations of the dielectric constant based on the amplitudes in a reference condition, a dielectric constant correction factor k = 0.87 should be adopted. During determinations of the dielectric constant based on the amplitudes at the temperature of −5 ◦C on the dry slab surface, a dielectric constant correction factor k = 0.97 should be adopted. During measurements at −5 ◦C and with a water film presence on the slab surface, a dielectric constant correction coe fficient k = 0.92 should be used. During measurements at −5 ◦C and with ice presence on the slab surface, a dielectric constant correction coe fficient k = 0.98 should be used. During measurements at −2 ◦C and with snow on the slab surface, a dielectric constant k = 1.02 correction coe fficient should be adopted. During measurements at −2 ◦C and with de-icing salt on the slab surface, a dielectric constant correction factor k = 0.55 should be used. As a result of applied correction factors, measurement errors of thickness determinations have been reduced, especially if there is de-icing salt on the surface: up to −4 ÷ +9% in the case of the surfaces covered with de-icing salt (see Table 12). In other cases, after applying correction factors, the thickness measurement error was up to ±4% in reference conditions, up to ±10% in the case of the surface below zero temperature, up to −7 ÷ +11% in the case of the surface covered with water film, up to −6 ÷ +9% in the case of icy surfaces, and up to −17 ÷ +8% in the case of the surfaces covered with a thin layer of snow. It is noted that the coe fficients have been calculated as the average of the coe fficients for 5 slabs with di fferent HMA wearing courses. To significantly reduce the error, a factor calculated for a specific HMA wearing course should be used.

Such coe fficients have been proposed for specific slab geometry (m1–m5), specific measurement conditions (wI–wVI), and selected antenna (1.0 GHz air-coupled). It is proposed to continue the research in order to build a base of dielectric constant values of various asphalt mixtures determined at di fferent temperature and humidity conditions based on the amplitude of the wave reflected from the asphalt layer and the corresponding correction factors, thus increasing the accuracy of the thickness determination of the layer by the GPR method and reducing the number of cores necessary to achieve the desired accuracy of thickness determination.
