Descriptive or normative: How does reservoir operations modeling influence hydrological simulations under climate change?
Descriptive or normative: How does reservoir operations modeling influence hydrological simulations under climate change?
Human activities have a strong influence on the hydrological cycle altering natural patterns of evapotranspiration, soil infiltration capacity, ice cover, groundwater distribution, and, ultimately, streamflow at different spatio-temporal scales. Yet, modelling human activities and the associated impacts are often given a secondary importance in hydrological models with respect to the high-fidelity characterization of natural processes, especially at the catchment scale. While this has little or no influence on model accuracy when modelling pristine watersheds, it remarkably deteriorates the model performance in river basins with a considerable anthropogenic footprint. Operations of water infrastructures within a watershed, e.g., a dam or a diversion dam, are commonly modelled based on observational data, when available. This reproduces to some extent the historical decisions, but might be inadequate to simulate operations outside of historical climate or socio-economic conditions. In this paper, we compare a descriptive approach traditionally adopted in hydrological models, where reservoir operations are determined by tracking the historical average release, and a normative approach, where the operations are dynamically conditioned upon the reservoir storage and can adapt to the climate and socio-economic conditions influencing the reservoir operations. We contrast these two approaches by assessing the reservoir dynamics and the impacts on the downstream river system across time scales, from daily to seasonal. We first discuss the accuracy of the two approaches in reproducing historical observations. Then, we explore their potential in anticipating the impacts of future reservoir operations when considering climate and socio-economic change scenarios, thus testing the approaches in decision making contexts increasingly altered with respect to the historical one. We critically present the advantages and disadvantages of either approach, thus contributing to clarify the importance of adopting an appropriate approach to model reservoir operations when reconstructing past dynamics or anticipating future dynamics of catchments impacted by human activities.
Climate change, Decision making, Human impacts, Hydrological modeling, Reservoir operations, Scenario-based analysis
Giudici, Federico
24e292ee-0b6a-4eb2-8a07-548fa5f52ab6
Anghileri, Daniela
611ecf6c-55d5-4e63-b051-53e2324a7698
Castelletti, Andrea
be719c8b-5599-42a5-8404-259074a780d6
Burlando, Paolo
5484fcec-b4d3-45e9-a72c-206ccbb5265f
1 April 2021
Giudici, Federico
24e292ee-0b6a-4eb2-8a07-548fa5f52ab6
Anghileri, Daniela
611ecf6c-55d5-4e63-b051-53e2324a7698
Castelletti, Andrea
be719c8b-5599-42a5-8404-259074a780d6
Burlando, Paolo
5484fcec-b4d3-45e9-a72c-206ccbb5265f
Giudici, Federico, Anghileri, Daniela, Castelletti, Andrea and Burlando, Paolo
(2021)
Descriptive or normative: How does reservoir operations modeling influence hydrological simulations under climate change?
Journal of Hydrology, 595, [125996].
(doi:10.1016/j.jhydrol.2021.125996).
Abstract
Human activities have a strong influence on the hydrological cycle altering natural patterns of evapotranspiration, soil infiltration capacity, ice cover, groundwater distribution, and, ultimately, streamflow at different spatio-temporal scales. Yet, modelling human activities and the associated impacts are often given a secondary importance in hydrological models with respect to the high-fidelity characterization of natural processes, especially at the catchment scale. While this has little or no influence on model accuracy when modelling pristine watersheds, it remarkably deteriorates the model performance in river basins with a considerable anthropogenic footprint. Operations of water infrastructures within a watershed, e.g., a dam or a diversion dam, are commonly modelled based on observational data, when available. This reproduces to some extent the historical decisions, but might be inadequate to simulate operations outside of historical climate or socio-economic conditions. In this paper, we compare a descriptive approach traditionally adopted in hydrological models, where reservoir operations are determined by tracking the historical average release, and a normative approach, where the operations are dynamically conditioned upon the reservoir storage and can adapt to the climate and socio-economic conditions influencing the reservoir operations. We contrast these two approaches by assessing the reservoir dynamics and the impacts on the downstream river system across time scales, from daily to seasonal. We first discuss the accuracy of the two approaches in reproducing historical observations. Then, we explore their potential in anticipating the impacts of future reservoir operations when considering climate and socio-economic change scenarios, thus testing the approaches in decision making contexts increasingly altered with respect to the historical one. We critically present the advantages and disadvantages of either approach, thus contributing to clarify the importance of adopting an appropriate approach to model reservoir operations when reconstructing past dynamics or anticipating future dynamics of catchments impacted by human activities.
Text
1-s2.0-S0022169421000433-main
- Version of Record
Restricted to Repository staff only
Request a copy
More information
Accepted/In Press date: 12 January 2021
e-pub ahead of print date: 1 February 2021
Published date: 1 April 2021
Keywords:
Climate change, Decision making, Human impacts, Hydrological modeling, Reservoir operations, Scenario-based analysis
Identifiers
Local EPrints ID: 448467
URI: http://eprints.soton.ac.uk/id/eprint/448467
ISSN: 0022-1694
PURE UUID: f8b76a2f-3ae6-4842-9795-5ecfbdcc19da
Catalogue record
Date deposited: 22 Apr 2021 16:47
Last modified: 17 Mar 2024 03:53
Export record
Altmetrics
Contributors
Author:
Federico Giudici
Author:
Andrea Castelletti
Author:
Paolo Burlando
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics