The purpose of this report is to investigate if thermal properties of dry permafrost can be determined by only using temperature measurements. Then see if a numerical solution to the heat conduction equation, with the assumption that the thermal properties are constant with depth and that the temperatures only varies sinusoidally with time, can reasonably predict temperatures with time and depth, using those derived properties. Thus giving an understanding of the thermal behavior in it. This is done in preparation for future Mars missions, because a good understanding of the physics in dry permafrost is needed to be able to make an educated guess of where to find life. One of the parameters that is used in the numerical solution, the thermal model, is the apparent thermal diffusivity which tells how fast/slow and large/small the changes in soil temperature are. This research utilizes two well-known methods that use amplitudes and phases of temperature measurements to determine the apparent thermal diffusivity. The focus is on analyzing and comparing four different methods to calculate the amplitudes and phases of temperature measurements taken at four different depths at Linneaus Terrace, Antarctica. The first two methods use the values of the maximum and minimum temperatures at the different depths to calculate their yearly amplitudes and phases. Of those two methods, one uses the actual data when determining the maximums and minimums while the other uses the same data that has first been smoothed by a box-car algorithm. The other two methods used are Fourier transform and least square best fit. Linneaus Terrace place is located in the high elevations valleys of the Antarctic Dry Valleys which is the only place on Earth with dry permafrost above ice-cemented ground. It is therefore a good analogue for Martian soil. The Fourier transform and least square best fit gave the most reasonable values for the data. They also gave the exact same values as each other, due to the use of a data set with a period of exactly one year. Just using the values of the maximum and minimum temperatures for the actual data was not a good way to find the yearly phases and amplitudes, due to fluctuations from other factors that have smaller periods than a year. Smoothing the data did give better values for the amplitudes, but the phases could not be properly determined due to the smoothing making all maximum temperatures end up in the first smoothed point. The temperature model was created using the yearly amplitudes, phases and apparent thermal diffusivity derived by the Fourier transform and the least square best fit. When comparing the result given by the model vs. the actual data at the four different depths (0, 0.17, 0.23 and 0.40m) at Linneaus Terrace, the model was found to reasonably predict the annual thermal behavior. More research needs to be made to verify the model on other data sets and with more coverage than only a year. This is only the first step in building a model for predicting at which depth and at what time a rover need to dig to find life on Mars. Moisture also needs to be added before it can be properly adapted to the Martian environment.