By Vasiliki Aspasiou
AbstractFor many geoscientists the collection, visualisation, storage and analysis of topographic data constitute an important factor in their studies. For that reason in many researches, the awareness of the quality of that data is extremely important. Due to the high demand of topographic data quality, many scientists have developed methodologies by using Digital Elevation Models (DEM); however, these methods are applied for a specific study area of specific research objective. In this study, an example database of field measurements have been used for assessing and gaining insights into topographical data quality.
The study area (Alora, Malaga Basin, S Spain) of this thesis is a region which is characterised by catenary soil formation. The objective of this research was the assessment of the dataset quality based on student observations and measurements comprising site-specific data. The topographical parameters under the scope of examination were elevation, slope, aspect, profile curvature and plan curvature. The DEM played a crucial role in the assessment of the field measurements quality, and was assumed as correct during all the phases of the research.
The first test that was implemented used the differences (?) between the values of the field measurements and the DEM values and derivatives and was applied per year, per geological mapping unit and per topographic parameter. For having an insight in data distribution, an Exploratory Data Analysis (EDA) was used. More specifically, histograms, scatterplots and Root Mean Square Error (RMSE) were used for further assessing the data distribution and quality.
The primary outcomes of this research indicate several coding errors in different parameters and varying over the observation years. Elevation and aspect were the parameters in which most of the coding errors occurred, 10% and 20 % respectively. While for slope and curvature, one and zero coding errors were identified respectively. Furthermore, generally the analysis of the distribution indicates some points as statistical outliers. The boxplots analysis demonstrated high annual variation. The students group that participated in 2009 can be characterised as “good” in terms of field measurement performance in all the topographic parameters. The group of 2004 proved to have lower performance in all the topographical parameters assessed. As for the results that were extracted by the test application per mapping unit, the findings showed that for Serpentinite lithology the elevation, slope and aspect differences are apparent to a greater degree of dispersion. Furthermore, the analysis of aspect per mapping unit indicated the presence of similarity trends between the metamorphic and the sedimentary mapping units.
Finally, for assessing the influence of topography in the elevation field measurements a boxplots analysis was applied only for the last two years, showing a higher accuracy elevation measurements according to the Global Positioning System (GPS) manufacturer. The results showed that high altitude and steep slope topography influence the elevation measurement accuracy. As for the observation in aspect, similar trends have been observed in N-S orientation.
Keywords: Topographic parameters; Field data; Álora; Catena; DEM; Mapping units; Data quality; accuracy; GIS; GPS