as function of the precipitation, evapotranspiration, surface runoff, sub-surface
runoff and deep percolation. The propagation process in the river channel is represent using a distributed Muskingum-Cunge model discretized each 1,000 m. More
details about these models can be found in Mazzoleni et al. [62].
The calibration of the hydrological model parameters was performed by AAWA
using an adaptation of the “SCE-UA” algorithm [73], considering the time series of
precipitation from 2000 to 2010, in order to minimise the root mean square error
between observed and simulated values of water level at PA (Vicenza) gauged
station. For the Muskingum-Cunge model, the only parameter that is calibrated in
this chapter is the Manning coefficient n, used to estimate the water level along the
river. The semi-distributed hydrological-hydraulic model in the Bacchiglione catchment is then validated considering the flood events that occurred in May 2013,
November 2014 and February 2016.
In order to apply data assimilation, both the hydrological and Muskingum models
are represented using the stochastic state-space form reported in the previous section.
In particular, for the Muskingum model, the approach proposed by Georgakakos
et al. [74] is used.
4 Model Updating Techniques
Operational forecast can be seen as combination of water models (e.g. hydrological
and hydrodynamic) and an updating module. In fact, in the last decades model
updating techniques have been intensively used within water system models
[75, 76], in order to reduce predictive uncertainty. The hydrological and
Legend
Sub-catchment A (Timonchio)
Sub-catchment B (Leogra)
Sub-catchment C (Orolo)
Inter-catchment 1 (Int C-1)
Inter-catchment 2 (Int C-2)
Inter-catchment 3 (Int C-3)
Inter-catchment 4 (Int C-4)
Internal boundary conditions
River network
StPh sensor
StSc sensor
N
S
E
W
0
3
6
12
18
Km
StPh-3
StSc-8
StSc-7
StSc-10
StSc-9
StPh-2
StSc-6
StSc-5
StSc-4
StSc-3
StSc-2
StSc-1
StPh-1
Fig. 3 Structure of the semi-distributed model for the Bacchiglione catchment and location of the
static physical (StPh) and social (StSc) sensors
216
M. Mazzoleni et al.
runoff and deep percolation. The propagation process in the river channel is represent using a distributed Muskingum-Cunge model discretized each 1,000 m. More
details about these models can be found in Mazzoleni et al. [62].
The calibration of the hydrological model parameters was performed by AAWA
using an adaptation of the “SCE-UA” algorithm [73], considering the time series of
precipitation from 2000 to 2010, in order to minimise the root mean square error
between observed and simulated values of water level at PA (Vicenza) gauged
station. For the Muskingum-Cunge model, the only parameter that is calibrated in
this chapter is the Manning coefficient n, used to estimate the water level along the
river. The semi-distributed hydrological-hydraulic model in the Bacchiglione catchment is then validated considering the flood events that occurred in May 2013,
November 2014 and February 2016.
In order to apply data assimilation, both the hydrological and Muskingum models
are represented using the stochastic state-space form reported in the previous section.
In particular, for the Muskingum model, the approach proposed by Georgakakos
et al. [74] is used.
4 Model Updating Techniques
Operational forecast can be seen as combination of water models (e.g. hydrological
and hydrodynamic) and an updating module. In fact, in the last decades model
updating techniques have been intensively used within water system models
[75, 76], in order to reduce predictive uncertainty. The hydrological and
Legend
Sub-catchment A (Timonchio)
Sub-catchment B (Leogra)
Sub-catchment C (Orolo)
Inter-catchment 1 (Int C-1)
Inter-catchment 2 (Int C-2)
Inter-catchment 3 (Int C-3)
Inter-catchment 4 (Int C-4)
Internal boundary conditions
River network
StPh sensor
StSc sensor
N
S
E
W
0
3
6
12
18
Km
StPh-3
StSc-8
StSc-7
StSc-10
StSc-9
StPh-2
StSc-6
StSc-5
StSc-4
StSc-3
StSc-2
StSc-1
StPh-1
Fig. 3 Structure of the semi-distributed model for the Bacchiglione catchment and location of the
static physical (StPh) and social (StSc) sensors
216
M. Mazzoleni et al.