PRODAM DSS

The regulation of an artificial reservoir is generally carried out in order to pursue different targets: agricultural, hydroelectric, ecological, recreational, protection from floods, etc All these objectives are often in conflict and their requirements in terms of water accumulation and release does not coincide neither in terms of time neither in terms of quantity. It is though necessary to find a compromise for reservoir management that on one side can minimize losses due to water deficits and on the other can reduce the entity and the number of the possible damages caused by floods.

PRODAM DSS is a reservoir management decision Support System that faces the problem of multi-purposes reservoir optimal management.

It uses the method of Stochastic Dynamic Programming and it helps the user defining the management plan of the water system considering the appraisal of the economic, ecological, environmental and social impact that it involves.

PRODAM DSS is especially dedicated to decision makers in charge of artificial reservoirs management who need a Decision Support System to guide them in optimizing water release, considering environmental and physical bindings. The functionalities of PRODAM include:

• Long-term optimization of water release (yearly scale)
  PRODAM DSS allows defining long-term water release strategy on the basis of available water and forecasted water income. The optimal amount of water to be allocated to final users is defined considering economic, ecological, environmental and social bindings.
• Daily optimization of water release strategy
  In order face to all those situations that do not fall in the stationary regulation of the reservoir, as for example the arrival of a flood, PRODAM software allows to determine, in real time, those modifications to long term management politics that become necessary in order to face non-stationary anomalous phenomena (for duration and/or entity).
• Hourly optimization of water release strategy
  For those cases in which a hourly regulation of the reservoir is necessary, for instance when a flood event is occurring, PRODAM allows to determine, in real time, those modifications to daily management politics that become necessary in order to face anomalous phenomena (for duration and/or entity).
• Inflow forecasting using Nearest Neighbours
  For forecasting daily inflow to the reservoir PRODAM uses Nearest Neighbors technique. In other words, on the basis of the last recorded daily inflows to the reservoir, the historical time series of daily inflows is searched for those situations of the past that are as similar as possible to the most recent data, according to an established metric. A number of "similar" time series is selected and on the basis of this sample a model is created to extrapolate future inflow values.

• Inflow forecast using real-time flood forecasting software EFFORTS
  Regarding the short term management rule, with a time horizon from one to twenty-four hours, PRODAM uses hourly forecasts of water inflow to the reservoir provided by a real-time flood forecasting system called EFFORTS.