Hydrologic modeling of an arctic watershed, Alaska

• Main Author: Schramm, I.
• Format: Thesis
• Language: unknown
• Published: 2005
• Subjects: Arctic; Nash; Sutcliffe; Climate change; permafrost; Alaska
• Online Access: https://epic.awi.de/id/eprint/12567/; https://epic.awi.de/id/eprint/12567/1/Schramm_05.pdf; https://hdl.handle.net/10013/epic.38924; https://hdl.handle.net/10013/epic.38924.d001

This study presents the application of the hydrological model TopoFlow to the Imnavait Creek watershed, a small arctic headwater basin in northern Alaska. This new process-based, spatially distributed model is executed for the years 2001 to 2003. The model is evaluated for its capability to reproduce the different components of the hydrological cycle. Simulations are done for different climate change scenarios to lend insight into the impacts of global change on hydrological processes. Imnavait Creek (~2 km²) is underlain by continuous permafrost and two features characterize the channel network: The stream is beaded, and numerous water tracks are distributed along the hillslopes. These facts, together with the constraint of the subsurface system to the shallow active layer, strongly influence the runoff-response to rain or snowmelt. Climatic conditions vary greatly during the years of this study, providing a good testing of model capabilities. Streamflow is the dominant form of basin water loss (64% of the water budget). In 2001, snowmelt runoff is the dominant runoff event, whereas in 2003, the summer runoff generated by continuous rainfall surpasses the melt discharge. A single and exceptionally high rainfall causes the dominant runoff event in 2002. Water loss due to evapotranspiration achieves considerable amounts, ranging from 28% to 57% of the water budget. Simulation results indicate that the model performs quantitatively well, and achieves best results in 2002. Measured and predicted cumulative discharges are in a good agreement. The different components of the water cycle are represented in the model, with refinements necessary in the qualitative reproduction of some sub-processes: Snow damming results in later melt discharge than modeled. Nash-Sutcliffe coefficients between 0.3 and 0.9 reveal that the model requires further refinement in the small-scale, short-term reproduction of storage-related processes. The deviations can be attributed to the facts that the beaded stream system, the spatial ...