HR: 14:40h
AN: H23L-05 [Abstracts]
TI: Risk and Reward: Adaptive Reservoir Operations using Seasonal Climate Forecasts
AU: Brown, C
EM: caseyb@iri.columbia.edu
AF: Columbia University - IRI, Lamont Doherty Earth Observatory, Palisades, NY 10964, United
States
AU: * Souza Filho, F A
EM: assis@iri.columbia.edu
AF: Columbia University - IRI, Lamont Doherty Earth Observatory, Palisades, NY 10964, United
States
AB:
A typical water manager's objectives consist of retaining enough water in the reservoir to meet urban demand
over the decision period, and releasing remaining water to agriculture in accordance with demand. The goal is
to release as much water as possible, but not so much that the urban demand is not met. As a result, there are
tradeoffs between releasing more water from the reservoir, and improving the yield. The decision hinges on the
assumption the water manager holds regarding the future inflows to the reservoir and the risk of an urban water
shortfall that the manager is willing to bear.
Here we explore the effects of the future inflows to the reservoir that are assumed in making reservoir release
decisions. With a conservative assumption of inflows, the reservoir releases will be limited by the continuous
need to conserve water for the future. This results in very high reservoir levels that are prepared to be drawn
down when the chosen drought scenario occurs. An alternative to these approaches is one that embraces the
known year to year variability in the probability of drought at a particular location. We term these approaches
"dynamic" risk management and provide an example using real interannual streamflow forecasts applied to
reservoir release decision making for Oros Reservoir, NE Brazil.
DE: 1869 Stochastic hydrology
DE: 1880 Water management (6334)
SC: Hydrology [H]
MN: 2007 Fall Meeting