| dc.description.abstract | Remotely Piloted Aircraft Systems (RPAS) have become an important tool in studying glaciers and their surfaces as they can generate high-resolution ortho-images and digital elevation models (DEM). It is important to monitor the surface of the Greenland Ice Sheet (GrIS) as it has become the largest cryospheric contributor to sea-level rise. Glacier surface energy balance models can provide insight in the mass loss rate and show that turbulent fluxes can form a major component. Aerodynamic roughness is an important parameter in these models, for which currently a fixed value is often used. This paper therefore analyses the possibilities of using RPAS-derived DEMs to quantify topographic roughness and its impact on turbulent energy fluxes on the GrIS. To answer this question, the following two unique methods were developed. First, the Moving Footprint (MF) method which produces a roughness estimate for a single location by moving a rotating footprint of varying size over the DEM. This footprint then extracts transects to which commonly used microtopographic algorithms (Lettau, Munro, and Nield) were applied. Second, the Fixed Grid (FG) method which produces a distribution of roughness values over the study area, by subdividing the study area in grid cells in which commonly used microtopographic algorithms (Lettau, Munro, and Nield) were applied. Our results show that both methods can produce realistic roughness estimates for a range of scales using certain algorithms. In the MF method, the Lettau algorithm performs well, while in the FG method, the Munro, Nield SdElev and Nield Max algorithms perform well. Moreover, we prove that the simulated sensible heat fluxes using roughness values derived using the MF method (Lettau, 5x40m footprint) or FG method (Munro, Nield SdElev, 5m grid size) are as accurate as simulated sensible heat fluxes using roughness values obtained by aerodynamic inversion of automatic weather station (AWS) data. The implication of these findings are that RPAS surveys are capable of producing realistic estimates of topographic roughness when using the methodology developed in this study, and thus contribute to the improvement of turbulent energy flux estimations for areas on the GrIS where no AWS are present. | |
