WWRP Example of hydrological modelling Slobodan Nickovic World

WWRP- Example of hydrological modelling Slobodan Nickovic World Weather Research Programme World Meteorological Organization, Geneva GAW with contributions in [B]: G. Pejanovic, M. Vujadinovic, V. Djurdjevic, B. Rajkovic (Belgrade University) 1

Example of hydrological modelling GAW WWRP- HYdrology PROgnostic Model integrated system DATASETS: HYDRO 1 k USGS topography HYDROSHED-500 m FAO soil texture data USGS land use data 2 Nickovic, et al (2010), HYPROM hydrology surface runoff prognostic model, Water Resour. Res.

WWRP- HYPROM - Full dynamic (FD) equation concept GAW Kinematic approximation neglects inertia forces !! 3

WWRP- Full dynamics (FD) vs. kinematics (KN) § § § GAW § § 4 FD model more accurate FD: friction slope term requires special treatment (Froude number >2) KN: simplifications avoid problem (Froude number <2) Most watershed models adapt KN approach KN cannot accurately represent large-scale, more inert processes

GAW WWRP- Kinematic approx Acceptable for flesh floods (where friction and gradient forces are in approx. balance) 5 Not appropriate for basins with slow flows Full dynamic system to be used

FD requires completely different numerical approach to resolve the Instability due to vanishing water heights! WWRP- Friction slope GAW Potential source of model instability when 6

WWRP- Friction slope numerics • Water depth is in the dominator of friction slope terms • Generates numerical instability when depth vanishes GAW • Usual approach - water depths to be above a threshold (not a physically-based approach) 7

WWRP- Friction slope numerics – new approach • Implicit time scheme applied • unconditionally stable method • convergent for • when then GAW • physically based method 8

GAW WWRP- Friction slope numerics – new approach Nickovic, et al (2010) 9

WWRP- Horizontal semi-staggered E grid used • the same as the NMM grid • problem with gravity wave short wave noise GAW • new method to resolve the problem – modification of the continuity equation 10

WWRP- Synthetic sink experiment with HYPROM GAW Without modification 11 With modification

WWRP- Advection numerics (Janjic, 1997) • conserves mass • no new extremes GAW • no negative values 12

WWRP- River routing • Mass conserving GAW • River –collector from surrounding points • Same numerics as for non -river points 13 River path: A-B-C-D-E-F

Savinja case – Oct/Nov 2000 WWRP- Torrential rain Landslides 7 deaths, quite a damage GAW Complex topography Kobold and Suselj, 2005 14 Savinja basin

WWRP- Savinja case – Oct/Nov 2000 Savinja basin GAW NMM T and V at 850 h. Pa, 1 Nov 2000 NMM acc. rainfall, 26 Oct – 6 Nov 2000 15

WWRP- Savinja case – Oct/Nov 2000 River discharge over period GAW 26/10 – 6/11 1990 16

WWRP- Moraca river experiment (Nov 2002 – May 2003) Snow melting case Seasonal-scale run GAW Problem with soil types 17

GAW WWRP- Sensitivity to soil type 18 parameter Clay Loam (09) Bedrock (15) sat. diffusivity 0. 113 x 10 -4 0. 136 x 10 -3 sat. conductivity 2. 45 x 10 -6 1. 41 x 10 -4 porosity 0. 465 0. 20 CH constant 8. 17 2. 79

water budget components 2008 accumulations WWRP- case 2008 GAW model vs. observations discharge 2008 19

GAW WWRP- climate studies Bojana river: Buna Bridge discharge 20