During the operation of a gas turbine, ingested contaminants present in the air form deposits on the surfaces of the turbine blades. These deposits grow over time, resulting in an increasingly rough surface. This gradual increase in roughness results in several negative consequences, among which is an increase in the rate of heat transfer to the blade which shortens blade life. This thesis presents research in which deposits were evolved on three different turbine blade coupons and their evolution was studied. A trend in roughness change over time was discovered. Also, an attempt was made to find the effect of the deposits on the heat transfer characteristics of a coupon surface. The deposits were formed using the BYU Turbine Accelerated Deposition Facility (TADF), which was used to simulate three months of deposition within a two hour test time. All three coupons underwent four cycles in the TADF: eight total hours of combustor testing—or one simulated year of deposition—with topological measurements being made on the coupon surface after every two hours (three simulated months) of testing. The data produced by the topological measurements were used with a CNC mill to machine scaled-up plastic models of the rough surfaces: four surfaces per model representing three, six, nine, and twelve simulated months of deposition. The models were placed in a wind tunnel where, following a period of thermal soaking at room temperature, they were suddenly exposed to a heated air stream. The thermal histories of the model were recorded with an infrared camera and were used to derive the heat transfer coefficient of each surface using the method developed by Shultz and Jones. The heat transfer coefficients are reported in the form of Stanton numbers to allow for the difference in thermal properties between the conditions and properties of the wind tunnel and its components and those of a real gas turbine. The Stanton numbers for the various surfaces were plotted versus the simulated gas turbine operational time. Additionally, several roughness correlations were used to predict the Stanton number for each surface, producing a probable Stanton number history for the coupon. The measured nondimensional heat transfer coefficients did not reach the magnitudes predicted by the correlations. This is most likely due to unexpected flow conditions inside the wind tunnel. Recommendations for future research are presented.