The value of hydrological sciences for operational applications and services. Günter Blöschl, TU Wien

Transcription

1 The value of hydrological sciences for operational applications and services Günter Blöschl, TU Wien Leibniz ( ) 7 May 2018, Geneva

2 The value of science for practice "He who loves practice without theory is like the sailor who boards ship without a rudder and compass and never knows where he may cast." Da Vinci ( ) new The value of science.. Four examples from my own research group

3 1 st example Science = Surface flow paths in karst areas - New mapping method (large scale) - Spatial information for distributed hydrol. models Reszler et al. (2018) HESS

4 - Flow path lengths from new mapping method - Testing against sink holes from Lidar Surface flow length (m) Reszler et al. (2018) HESS

5 Modelled spatial patterns of surface runoff based on flow paths 4. Aug :30 Messstelle Karst Messstelle 12. Aug :00 Karst km Reszler et al. (2018)

6 Application = Land management to protect karst springs (Vienna water supply) - Surface runoff may enhance pollution Value = Prioritisation of protection zones, management options

7 2 nd example Science = New methods of ensemble streamflow forecasts - Soil moisture updating based on streamflow - Ensemble spread consistent with forecasting uncertainty Wien Precipitation Streamflow Forecast points

8 Ensemble forecasts Streamflow (m³/s) Cumul. Precip. (mm) Time forecast is issued Precipitation Ensemble spread Zwettl/Kamp 620km² Observation Deterministic forecast Ensemble member 9 July 2005 Time (hrs) Blöschl et al. (2005)

9 Forecast error (s.dev.) Ensuring consistency of ensemble spread with forecast uncertainty of streamflow Total forecast error Forecast error due to precipitation forecast Small catchments Large catchments Ensemble spread (s. dev.) Ensemble spread (s. dev.) 48 hr forecasts for tributaries to Danube Errors and spread scaled by mean streamflow Nester et al. (2011) WRR

10 Streamflow (m³/s) Application = Flood warning at the Danube by governments; Cargo optimisation low flows (Viadonau) Value = More reliable assessment of uncertainties June 2013 Forecast m³/s Observed m³/s

11 Peak runoff (m³/s) Ruonff 3 rd Example Science = flood frequency hydrology Combining process understanding + statistics Precipitation Spatial information Causal information Temporal information Flood data Return period Merz und Blöschl (2008) WRR

12 Causal expansion of information e.g. Frequency of flood process types Snow melt Flash c) floods Rain-on-snow Synopt. floods Events Catchments >100km² Frequency Merz und Blöschl (2004) WRR

13 Causal expansion of information Eg. timing of floods within the year to understand changes Stations, Blöschl et al. (2017) SCIENCE

14 Application = Flood risk mapping Austria (HORA) (Ministry, states, insurance industry, general public) Value = More reliable risk estimates under change Flood prone areas Projekt HORA -

15 4 th example Science = propagation of flood probabilities through catchments - Robust spatial stochastic precipitation modelling - Regional Monte Carlo Simulations Bavaria Innsbruck Switzerland Tirol 25 km Urban Forest Grassland Cropland Grassland Rocks Glacier

16 Stochastic precipitation model Simulated Observed mm/d Salinas et al. (2018)

17 Abfluss (m³/s) Monte Carlo Simulations (spatially distributed rainfall-runoff) - Testing return periods of flood runoff spatially - Confluence of floods how do return periods combine? Oberaudorf Zeit (h) Nester et al. (2018)

18 WLV Tirol S. Jud Application = Strategic flood risk management at the Inn (Tirol, Bavaria) Value = Understanding effectiveness of retention reservoirs in headwater catchments for main stream Effective Innsbruck not effective Effectiveness 25 km Nester et al. (2018)

19 Good science leads to better applications, which lead to better science Science.. published in recent literature 1. Surface flow paths in karst areas 2. New methods of ensemble streamflow forecasts 3. Flood frequency hydrology 4. How do flood probabilities propagate through catchments? Applications and services.. direct societal value Land management to protect karst springs (Vienna water supply) Flood warning at the Danube; Cargo optimisation (Viadonau) Flood risk mapping (HORA), including change Strategic flood risk management at the Inn (Tirol, Bavaria) Applications conducive to better science Understanding karst spring behaviour Testing assimilation of Earth Observation data (snow, soil moisture) High quality data set for testing new methods Regionalisation of stochastic precipitation model