pact 
sing 
e to 
ated 
298) 
| the 
ibly 
ntly 
e of 
rge 
vith 
stic 
fail 
lon. 
vith 
not 
  
Castaldini, Doriano 
  
event. Furthermore, a breach location or spillage zone has to be specified, for which records on previous flooding events 
is used. Temporal input for the model consists of precipitation and discharge records (5 minute or hourly intervals). 
Spillage from the channel is calculated as a function of channel discharge. 
Flood routing (the process of tracing water movement through determination of time and magnitude of flow in the 
watershed; Chow et al., 1988) is performed treating flow of water as kinematic wave motion, where motion of water is 
described principally by the equation of continuity. Water moves according to the steepest gradient from one element to 
another. For each time interval (5min or Ihr) the quantity of water in each grid-cell (25 or 50 m^) is updated as a 
function of input and output. Three problems arise from the use of a drainage network for routing purposes: 1) Flow 
restrictions: water can only move as far as the (diagonal) length of 1 grid element. To resolve this problems the pixel- 
size and used time interval can be reduced (increasing computational expense); 2) Micro-relief: the presence of 
infrastructures does not become apparent in the digital elevation model (due to its vertical resolution). To resolve this 
problem additional relief is added where infrastructure is present; 3) Local drain direction is fixed: Water can only 
discharge to 1 specific downstream pixel during the entire simulation. Therefore the digital elevation model is 
recalculated after an arbitrary time interval on the basis of the actual elevation and the quantity of water present in a grid 
cell. 
42 Model output 
  
  
For every interval a map is 
created depicting the quantity of 
water in a grid-cell. In these 
maps areas with urbanisation can 
be identified. Water-level in time 
can be reported as a hydrograph 
at any location in the area. 
In figure 5 two simulations are 
displayed, the first with one of 
  
  
  
three alternative routes for the 
railway and a second for the 
present situation. 
  
simulated flooded area for presence of railway 
" boundary of Castelfranco Emilia Municipality 
^w. alternative high velocity raiheay track 
  
  
5 CONCLUSIONS 
From the study of meteorological 
data it became clear that on the 
basis of merely precipitation no 
predictions can be made with 
respect to the occurrence of a 
flood event. Given the 
difficulties in understanding and 
thus modelling the mechanisms 
  
huilt up area 
flooded area September 1973 
w«^ boundary of Castelfranco Emilia Municipality 
= Panaro river | à 
€ simulated flooded area for present situation 
boundary of Castelfranco Emilia Municipality 
      
  
  
  
  
  
  
that provoke an event, it is 
thought to be more fruitful to 
make predictions with respect to 
consequences of what-if 
scenarios, rather than predictions with respect to occurrence. Or, as Kundzewicz (1999) puts it: “it is necessary to be 
aware of the possibility of floods... living with floods seems more sustainable than a hopeless striving to combat floods”. 
In this light, a distributed dynamic model is constructed which allows for the calculation of flood scenarios, which can 
be fed by historical or hypothetical precipitation and discharge records. 
The inclusion of alternative routes for the proposed railway allows for the evaluation and comparison of the spatial 
effects of a flood event. Critical points in the model are calibration and the calculation of soil saturation prior to the 
event. 
Figure 5. Presence of water on surface for the 1973 flood (left), and a simulation 
for one of the proposed railway routes (upper right) and the same simulation for 
the current situation (lower right). 
ACKNOWLEDGMENTS 
The financial support for the study was provided by the European project: *A European Research Network for the 
Application of Geomorphology and Environmental Impact Assessment to Transportation Systems (GETS)", (contract 
ERBFMRXCT970162), Local Co-ordinator of the Modena and Reggio Emilia University, Prof. M. Panizza. 
  
International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B7. Amsterdam 2000. 233