Modeling of the Lake Pareloup ecosystem with the ASTER and MELODIA models

• Published in: Hydroecologie appliquee. Paris Vol. 6; no. 1-2; pp. 369 - 426
• Main Authors: Salencon, MJ, Thebault, J M
• Format: Journal Article
• Language: French
• Published: 01.01.1994
• Subjects: Freshwater; France, Midi-Pyrenees, Pareloup L
• ISSN: 1147-9213

A four-year research project conducted at Pareloup lake, an Electricite de France hydroelectric reservoir, improved understanding of the dynamics of the ecosystem in both biological and hydrodynamic terms. The simulation models developed during the project enabled the proposal of a tool to aid in water resource management. The first phase was development of the ASTER model. This is a biological model (two phytoplankton groups, three zooplankton groups, detritus, PO sub(4) and SiO sub(2)) with a simple physical structure (two-layer model). The sizeable data base available enabled calibration of the model and ensuring that it correctly simulated seasonal evolution of each of the variables over a five-year period (1983-1987). Thanks to this model, the preponderant mechanisms in the evolution of biological variables were highlighted. The great sensitivity of the ecosystem to hydrodynamic factors prompted us to undertake the second phase, which was development of the MELODIA model. This is a reservoir ecosystem management model obtained by coupling the two-layer biological model (ASTER) with a vertical hydrodynamic and thermal model (EOLE) which simulates the evolution in thermal stratification and takes into account energy exchanges across the air-water interface, as well as inflow and outflow. This model enabled simulation on a one-day scale of the vertical dynamics of the ecosystem for the five years of measurements in the data base. Comparison of the calculated and measured profiles for each of the variables indicates quite good representation of the ecosystem, both in terms of vertical dynamics and in terms of seasonal evolution. Furthermore, year-to-year fluctuations in the ecosystem are well represented, and simulations showed the extent to which the ecosystem is sensitive to the hydrodynamic structure. The model was then used to study different reservoir management modes. Applied to a longer period of time, 1976-1987, it pointed up the year-to-year fluctuations due to hydrometeorological conditions and identified periods in which the risk of eutrophication is great.