id, 
ea 
or 
nde 
ing 
the 
ted 
ter 
tra 
Castaldini, Doriano 
  
The main and auxiliary storage basin is completely bordered by embankments about 6 km long and 4-5 m high that were 
constructed to increase the capacity of the storage area. 
d) a downstream barrier: a transverse structure constructed downstream from the regulating dam and which serves the 
purpose of maintaining the river bed level, thus ensuring protection against bed erosion. 
e) a selective barrier: a structure constructed upstream from the regulating dam which consists of a transverse net which 
serves to prevent large plants and trees from blocking or interrupting the discharge from the apertures in the regulating 
structure. 
The flow regulation system has not yet been tested as to its efficiency under the flood conditions for which it was built, 
anyhow since this hydraulic structures have been in operation no floods have occurred. 
3.2 Meteorological data 
On the basis of the database it was not possible to forecast floods on the basis of merely precipitation. The three 
recorded floods are concentrated around the months September and November (Nov. 1966, Sept. 1972 and Sept. 1973) 
(Fig. 3). The question is if it is possible to identify a 
trigger mechanism (e.g. exceedence of a threshold 
quantity of precipitation or critical stage) to which 
the 3 more recent floods comply. This does not seem 
the case: the September events stand apart from the 
one that took place in November. A good illustration 
is the fact that e.g. in 1966 a flood occurred under 
meteorological circumstances which were far from 
exceptional. It appears however that soil moisture 
conditions play a role in the process. The two floods 
that occurred at the beginning of the wet season 
came after extreme amounts of precipitation, in some 
cases the highest amounts ever observed. The other 
flood (November 1966) occurred after much less 
intense precipitation, but were preceded by wet 
periods. Therefore (partial) saturation of soil storage 
capacity throughout the study area, would favour 
  
runoff. Antecedent soil moisture conditions could, in Figure 3. Brosch in the embankments of the Panaro river 
combination with precipitation, play an important ^j occasion of the flood occurred on 25-09-1973 (Moratti 
role in the triggering of an event. Further study of ^ $ Pellegrini, 1977). 
soil moisture could aid confirmation of this surmise. 
There is however a fairly linear relationship between the amount of precipitation that falls immediately prior to the event 
and the magnitude of peak discharge. 
Flood forecasting on the basis of (daily) rain intensities along with information on the saturation of soils (e.g. the freatic 
water-level) will probably give more satisfying results, since the events that occurred directly after the dry season (Sept. 
72 and 73) were accompanied by much larger quantities of precipitation than other event, which was preceded by 
relatively wet months. 
4 MODELLING OF FLOOD PROPAGATION 
In general flood control strategies range from correlation of flood potential by means of simple descriptive indices to 
more sophisticated analyses of flow and sediment transport (Dunne, 1988). These two extremes coincide with the two 
main types of flood forecasting: 1) on the basis of frequency relationships; and 2) on the basis of dynamic models. In the 
first statements are made on the basis of statistical relationships (e.g. Meigh ef al., 1997 (recurrence intervals); Jiang, 
1998 (probability of flood release due to overtopping)). In the second usually conceptually physically-based runoff 
models (also conceptual rainfall-runoff models, also CRRMs; Franchini & Galeati, 1997) are used that transform 
precipitation input into a quantitative description of stream response (E.g. Bentura & Michel, 1997; Gôppert et al. 
1998; Romaniwicz & Beven, 1998; Peschke, 1998; Chang & Hwang, 1999; Estrela Monreal, 1999). These forecast 
models aim at providing forecasts for a discharge hydrograph. Samuels (1998) notes the importance of hydrological 
  
International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B7. Amsterdam 2000. 231