Since reservoir inception, neither discharge nor
water levels have been systematically monitored.
A modelling approach is used for a retrospective
performance and sustainability assessment of
these four reservoirs.
2 Materials and Methods
2.1 Model Description
Hydrological models can vary in complexity,
depending on the number of processes represented and the purpose of the modelling. Complex models are needed to understand complex
feedbacks and interactions among different processes, but typically require many input data and
extensive parameterization. Much simpler
‘bucket’ models only represent the most
important processes (Zhang et al. 2002). They
are water balance models that conceptualize the
system of interest as a ‘bucket’, being filled up
by rainfall and emptied by evapotranspiration
and infiltration. When the bucket is full, extra
water is assumed to overflow. Although their
simplicity restricts their versatility and generalizability, bucket models can provide useful
insights into the core behaviour of a system
(ibid.).
In this study, a simple lumped ‘bucket’ model
calculates reservoir water volumes as a function
of inputs and losses to the reservoir (Fig. 2).
Inputs are due to surface runoff from precipitation over a reservoir’s catchment, losses are due
to infiltration, evaporation, extraction and overflow. Subsurface hydrology, vegetation hydrology and spatial heterogeneity are not accounted
for in this simple model.
Fig. 1 Study area, with land use pattern and reservoirs
1A, 1B, 2 and 3. An interconnection allows any overflow
from reservoir 1A to run into reservoir 1B (thick blue line,
schematic). Yellow line marks farm boundary. Thin blue
lines indicate pre-reservoir ephemeral channel drainage
network. Inset shows location of the study area in Spain.
(Background image source Google Earth, 37°23′08.5″N
7°01′19.7″W, April 2013)
62
I. Fiebrig and M. Van De Wiel
water levels have been systematically monitored.
A modelling approach is used for a retrospective
performance and sustainability assessment of
these four reservoirs.
2 Materials and Methods
2.1 Model Description
Hydrological models can vary in complexity,
depending on the number of processes represented and the purpose of the modelling. Complex models are needed to understand complex
feedbacks and interactions among different processes, but typically require many input data and
extensive parameterization. Much simpler
‘bucket’ models only represent the most
important processes (Zhang et al. 2002). They
are water balance models that conceptualize the
system of interest as a ‘bucket’, being filled up
by rainfall and emptied by evapotranspiration
and infiltration. When the bucket is full, extra
water is assumed to overflow. Although their
simplicity restricts their versatility and generalizability, bucket models can provide useful
insights into the core behaviour of a system
(ibid.).
In this study, a simple lumped ‘bucket’ model
calculates reservoir water volumes as a function
of inputs and losses to the reservoir (Fig. 2).
Inputs are due to surface runoff from precipitation over a reservoir’s catchment, losses are due
to infiltration, evaporation, extraction and overflow. Subsurface hydrology, vegetation hydrology and spatial heterogeneity are not accounted
for in this simple model.
Fig. 1 Study area, with land use pattern and reservoirs
1A, 1B, 2 and 3. An interconnection allows any overflow
from reservoir 1A to run into reservoir 1B (thick blue line,
schematic). Yellow line marks farm boundary. Thin blue
lines indicate pre-reservoir ephemeral channel drainage
network. Inset shows location of the study area in Spain.
(Background image source Google Earth, 37°23′08.5″N
7°01′19.7″W, April 2013)
62
I. Fiebrig and M. Van De Wiel