1. RCP4.5 and RCP8.5 runoff regime changes in a regional-scale glacierized catchment in the Austrian Alps
Florian Hanzer (1,2), Kristian Förster (1,2), Thomas Marke (2), and Ulrich Strasser (2)
alpS Centre for Climate Change Adaptation, Innsbruck, Austria
Institute of Geography, University of Innsbruck, Austria
2nd ANNUAL Workshop of GEWEX’s INARCH - “French Alps” – 19 October 2016

2. Aim
Development of a scenario-capable model system for the quantification of future changes in annual and seasonal discharge in a snow- and icemelt-dominated catchment of regional scale
Application of a comprehensive validation method
Analysis of the regional patterns of the current and future snow- and icemelt contribution to discharge volume
Learn what should be further improved  !

3. 1760 m a.s.l. – 3770 m a.s.l. 558 km2, 24 % glac. cov.

4. The model
(Brandwirt, Gosau)

5. Model setup
Initial ice conditions derived for 1997 from the 2nd Austrian Glacier Inventory
Glacier retreat parameterization (Dh) by Huss et al. (2010)
Layers for old/new snow, firn, ice with parameterization for cold content and liquid water content
Lateral snow transport by means of the topographic openness (50m/5000m length scales, mass-conserving redistribution)
Snow-canopy interaction with interception, sublimation losses, melt unload and separate beneath-canopy snow cover simulation
Modules for evapotranspiration and runoff generation
Multitlevel spatio-temporal validation procedure based on the observation scale of of support, spacing and extent as well as information type (validogram)

6. Wind-induced snow transport The snow redistribution factor
Negative openness for length scales 50 m (left) and 5000 m (center), and respective snow redistribution factor (right). Hanzer et al. (2016).

7. Model setup
Initial ice conditions derived for 1997 from the 2nd Austrian Glacier Inventory
Glacier retreat parameterization (Dh) by Huss et al. (2010)
Layers for old/new snow, firn, ice with parameterization for cold content and liquid water content
Lateral snow transport by means of the topographic openness (50m/5000m length scales, mass-conserving redistribution)
Snow-canopy interaction with interception, sublimation losses, melt unload and separate beneath-canopy snow cover simulation
Modules for evapotranspiration and runoff generation
Multitlevel spatio-temporal validation procedure based on the observation scale of of support, spacing and extent as well as information type (validogram)

8. Validogram: the observation scale
Support = integration volume or time of a single sample
Spacing = distance or time between individual samples
Extent = total coverage in space or time of the entire data set
Observation scale of a set of observations in space and time (after Blöschl & Sivapalan 1995). Hanzer et al. (2016).

9. Validogram
Validogram axes for the visualisation of the observation scale of the validation data sets. Hanzer et al. (2016).

10. Validogram
Validogram with observation scale of the used validation data sets (top), as well as their information type (binary/continuous, spatial/temporal support, spacing, extent) (bottom). Hanzer et al. (2016).

11. Modelling of streamflow components
Observed vs modelled streamflow (with components) for the gauge Gepatschalm ( m a.s.l.). Hanzer et al. (2016).

12. Quality check: seasonal/longterm validation –
comparison with observed mass balances
Glacier inventories 1997 and Model
Observed vs modelled elevation changes of glacier surfaces (1997 – 2006, Ötztal Alps/Austria). Hanzer et al. (2016).

13. Scenario modelling: meteorological forcing
Selection of moderate, wet and warm realizations from EURO-CORDEX rcp4.5 and rcp8.5
Statistical downscaling and bias correction with quantile mapping
Disaggregation to 3-hourly values using statistical patterns derived from hourly observations with MELODIST (Förster et al. 2016)

14. Scenario modelling: future glacier evolution
Visualisation of the glacier evolution model: Ice thickness (Ötztal Alps/Austria), initialized with ice thickness 1997 (Austrian glacier cataster), temperature change (for Austria) °C/year, glacier evolution model after Huss (2010).

15. Scenario modelling: future glacier evolution
Glacier volume change in the Ötztal Alps/Austria until 2050 and 2100 for rcp4.5 and rcp8.5; Hanzer et al. (2016).

16. Scenario modelling: future runoff regime
Change of the streamflow regime of the Rofenache (Vent, 1891 m a.s.l.) for rcp4.5 and rcp8.5. Hanzer et al. (2016).

17. Conclusions and outlook
RCP4.5 and RCP8.5 simulations both indicate severe decrease in glacier volume until 2100
Runoff regime changes from glacial to nivo-glacial and nival, respectively
Significant uncertainty through choice of GCM/RCM realization, and surface process representations ...
To do:
Better climate scenario data (and not „if swe > 10 m then swe = 10 m“)
Better downscaling ... (bidirectional coupling RCM-surface)
Finding the right models.. ?!
Hanzer, F., Helfricht, K., Marke, T. and Strasser, U. (2016): Multi-level spatiotemporal validation of snow/ice mass balance and runoff modeling in glacierized catchments. The Cryosphere, 10, 1859–1881,