Digitizing and Mapping Channel Heads Across Various Forested Regions in Sweden Using a High-Resolution DEM : A Regional Analysis of How Landscape Properties Influencing Channel Head Initiation

University essay from Uppsala universitet/Institutionen för geovetenskaper

Author: Amanda Jarefjäll; [2021]

Keywords: ;

Abstract: Acting as a hydrological link between the terrestrial landscape and the larger downstream river network, channel heads and their immediate headwaters are essential geomorphological features in our landscape. Not only due to their ability to transport water and sediment and cycling of nutrients, but as an important component for ecosystem productivity and aquatic habitats. Yet, many of our low order streams remain poorly mapped which consequently make development of management policies and protection of them more difficult. A possibility to derive more accurate stream network maps may be achieved by using high resolution digital elevation data (DEM) and hydrological models. A central part in stream network modelling is to develop reliable methods to predict the start of a channel, i.e. the channel heads. Different landscape variables influence the formation and maintenance of channel heads, stressing the need to study their controls in different physiographic regions. Factors influencing channel head formation in the Swedish landscape are yet still relatively unexplored. Investigating the regional and temporal controls on in this landscape are especially interesting since it has a long history of human alterations such as digging of extensive drainage networks, deepening, and straightening of natural stream channels.     The objective of this study is to explore channel head distribution and formation in 12 different forested regions in Sweden using statistical - and GIS analysis. By digitally tracing channel networks upstream using ArcGIS software and a high-resolution DEM (0.5 x 0.5 m), a total of 394 channel heads have been identified and digitized. By calculating the channel heads contributing areas, i.e. the drainage area threshold needed for channels to form, this study have examined regional variations in the size of the contributing area and whether the variability could be attributed to local and average upslope topographic-, soil and hydrological indices. 42 % of the digitized channel heads were identified as being subjected to man-made alterations in some way, remaining were classified as naturally formed channel heads. The contributing area to the channel heads reveled large variability ranging from 0.1 to 58.0 ha for natural to 0.04 to 193.6 ha for more altered channel heads. No evident relationship between size of contributing area and type of Quaternary deposit could be found. Highest median was found in channel heads formed on fine sediment, such as clay, (8.9 ha) however no statistically significant difference in median between different deposits could be confirmed. Channel heads showed only insignificant to weak correlations to the extracted landscape variables, such as the catchment slope and soil moisture index. This could partly be explained by the relatively gentle topography in the investigated regions. The results within this study have highlighted the difficulties to apply traditional methods using topographic indices when delineating and predicting the start of the stream channels in forested and cultivated regions in Sweden. Yet, it has explored the possibility of using high resolution DEM to map and analyze small-scaled landscape features such as channel heads. 

  AT THIS PAGE YOU CAN DOWNLOAD THE WHOLE ESSAY. (follow the link to the next page)