Statistical criteria

For the quantitative validation of the simulation results, the Nash-Sutcliffe efficiency (Nash and Sutcliffe 1970) and the total runoff amount difference between measurement and simulation is calculated. The Nash-Sutcliffe criterion for the daily discharges is given in Table 4.3 for the three subcatchments modeled.

The table shows that the 8-km results (Nash-Sutcliffe efficiency = 0.73 for La Clapiere) are clearly improved by consideration of the subgrid topography (Nash-Sutcliffe efficiency = 0.77 for La Clapiere) and also slightly improved by application of the forest climate model (Nash-Sutcliffe efficiency = 0.75 for La Clapiere). The combination of both parameterisations leads to results comparable with the 1-km resolution reference, for Briancon the Nash-Sutcliffe criterion is even higher (Nash-Sutcliffe efficiency = 0.63) than the reference (0.61).

Table 4.3 Nash-Sutcliffe efficiency for 14 years of daily discharge (1981/82 to 1994/95) for the gauging stations Briancon, L'Argentiere and La Clapiere

Briancon L'Argentiere La Clapiere

Table 4.3 Nash-Sutcliffe efficiency for 14 years of daily discharge (1981/82 to 1994/95) for the gauging stations Briancon, L'Argentiere and La Clapiere

Briancon L'Argentiere La Clapiere

1km

0.61

0.73

0.79

8km

0.13

0.60

0.73

8 km/Topography

0.59

0.66

0.77

8 km/Forest

0.29

0.66

0.75

8 km/Topography

0.63

0.69

0.78

and Forest and Forest

Table 4.4 Total runoff amount difference in percentage (modeled - observed) for 14 years of daily discharge (1981/82 to 1994/95) for the gauging stations Briancon, L'Argentiere and La Clapiere

Briancon L'Argentiere La Clapiere

Table 4.4 Total runoff amount difference in percentage (modeled - observed) for 14 years of daily discharge (1981/82 to 1994/95) for the gauging stations Briancon, L'Argentiere and La Clapiere

Briancon L'Argentiere La Clapiere

1km

0.01

-3.58

-4.56

8km

-1.79

-3.39

-5.60

8 km/Topography

-2.48

-3.65

-6.22

8 km/Forest

-5.88

-7.41

-9.68

8 km/Topography

-6.78

-7.87

-10.56

and forest and forest

The total runoff amount difference is an indicator if the simulated partitioning of precipitation into runoff and evaporation is correctly represented by the model. As can be seen in Table 4.4, the mean total runoff amount difference is negative for all stations (except for Briancon and the 1-km reference experiment) and increases with growing catchment size (i.e., for the subgrid parameterisation of topography for Briancon -2.5%, for Argentiere -3.7% and for La Clapiere -6.2%); it also increases with the forest experiment (-5.9, -7.4 and -9.7%) and is largest for the combined experiment (-6.8, -7.9 and -10.6%). This reduction in total runoff is due to the interception losses as parameterised with the forest climate model. However, part of the reduction in modeled runoff with the subgrid parameterisation of topography is due to the reduction in modeled precipitation inputs evident in Table 4.5, attributable to the interpolation of meteorological variables.

If one looks at the interannual comparison of the statistical criteria (Figure 4.6), the following can be observed.

• The differences between the Nash-Sutcliffe efficiencies for the three experiments decrease with growing catchment size because of increasing temporal duration and spatial variation of snowmelt.

Briangon

Briangon

1 km 8 km/Topography -»»-8 km/Topography and forest -8 km -.»-8 km/Forest

1 km 8 km/Topography -»»-8 km/Topography and forest -8 km -.»-8 km/Forest

Briangon

Briangon

-8 kkm 8 kkÄraphy 8 km/Topography and forest

-8 kkm 8 kkÄraphy 8 km/Topography and forest

Argentiere

8 km/Topography -«-8 km/Forest 8 km/Topography and forest

8 km/Topography -«-8 km/Forest 8 km/Topography and forest

Argentiere

1 km 8 km/Topography -■-8 km/Topography and forest -8 km -«-8 km/Forest

La Clapiere

La Clapiere

8 km/Topography -«-8 km/Topography and forest 8 km/Forest

8 km/Topography -«-8 km/Topography and forest 8 km/Forest

^^VVVVVVVV*

La Clapiere

La Clapiere

8 km/Topography -—8 km/Forest -—8 km/Topography and forest

8 km/Topography -—8 km/Forest -—8 km/Topography and forest

A^VVVVVVVVVVVV"

0 0

Post a comment