Multilevel spatiotemporal validation of snow/ice mass balance and runoff modeling in glacierized catchments

Abstract

Abstract. In this study, the fully distributed, physically based hydroclimatological model AMUNDSEN is set up for catchments in the highly glacierized Ötztal Alps (Austria, 558 km2 in total). The model is applied for the period 1997–2013, using a spatial resolution of 50 m and a temporal resolution of 1 h. A novel parameterization for lateral snow redistribution based on topographic openness is presented to account for the highly heterogeneous snow accumulation patterns in the complex topography of the study region. Multilevel spatiotemporal validation is introduced as a systematic, independent, complete, and redundant validation procedure based on the observation scale of temporal and spatial support, spacing, and extent. This new approach is demonstrated using a comprehensive set of eight independent validation sources: (i) mean areal precipitation over the period 1997–2006 derived by conserving mass in the closure of the water balance, (ii) time series of snow depth recordings at the plot scale, (iii–iv) multitemporal snow extent maps derived from Landsat and MODIS satellite data products, (v) the snow accumulation distribution for the winter season 2010/2011 derived from airborne laser scanning data, (vi) specific surface mass balances for three glaciers in the study area, (vii) spatially distributed glacier surface elevation changes for the entire area over the period 1997–2006, and (viii) runoff recordings for several subcatchments. The results indicate a high overall model skill and especially demonstrate the benefit of the new validation approach. The method can serve as guideline for systematically validating the coupled components in integrated snow-hydrological and glacio-hydrological models.

Projected cryospheric and hydrological impacts of 21st century climate change in the Ötztal Alps (Austria) simulated using a physically based approach

“…AMUNDSEN has specifically been designed as a scenario-capable model for the application in high-mountain regions, and has been set up and extensively validated for historical conditions in the study site in a recent study (Hanzer et al,5 2016). In the following, the most important model components are briefly discussed -for a more detailed model description, we refer to, e. g., Hanzer et al (2014Hanzer et al ( , 2016, Marke et al (2015), Pellicciotti et al (2005), Strasser (2004Strasser ( , 2008, .…”

Glacio-hydrological melt and runoff modelling: a limits of acceptability framework for model selection

“…These data have been archived since 2000 and consist of daily 500 m gridded maps of snow cover extent with values ranging between 0 and 1 which relate to the proportion of the ground that is snow covered. While they do not provide a direct measurement of snow mass balance, they have shown to be a useful data source for evaluating the performance of GHMs (Hanzer et al, 2016;Finger et al, 2015). The quality of the data in high latitude regions such as Iceland are variable due to the need for good light and little or no cloud cover.…”

“…Arguably, the equations that govern the routing (transport) of runoff are most important in relation to river flow predictions in glaciated river basins, as storage characteristics of ice and snow strongly influence river flow regimes over a range of time-scales (Jansson et al, 2003). The concept of linear reservoirs is the most widely adopted simplified approach for runoff-routing in glaciated basins (Zhang et al, 2015;Hanzer et al, 2016;Gao et al, 2017). A linear reservoir lumps all of 25 the interacting, non-linear and non-stationary components of water transmission within a pre-defined area (e.g.…”

“…This is partly because the concept lends itself to structural modifications that can represent different glacio-hydrological systems. Hanzer et al (2016) hypothesised that the snow pack, firn layer, glacier ice and the region free 30 from ice all exhibit unique runoff-discharge responses and advocate the use of four linear reservoirs in parallel to distinguish between these units. However, simpler structural configurations using only two linear reservoirs in parallel to route meltwater through the snowpack and ice separately (Hannah and Gurnell, 2001) or even a single linear reservoir to route all rainfall and melt runoff simultaneously (Boscarello et al, 2014) can accurately reproduce river discharge time-series.…”

“…For this purpose, signatures need not be derived just from river discharge data, but should also be taken from other observation sources such as ice melt and snow coverage as these have shown to be useful for evaluating the consistency of GHMs across different aspects of glacio-hydrological systems (Finger et al, 2015;Hanzer et al, 2016;Mayr et al, 2013;Finger et al, 2011). By doing so, this framework could facilitate better predictions of river flow regime changes in glaciated river basins; firstly by helping to diagnose 25 deficiencies in GHM structures that require improvement, and secondly, by objectively selecting the range of acceptable model structures and parameterisations so that prediction uncertainty can be better constrained.…”