Citation:

Bousquin, J., W. Fisher, J. Carriger, AND E. Huertas. A Bayesian Belief Network Approach to Explore Alternative Decisions for Sediment Control and water Storage Capacity at Lago Lucchetti, Puerto Rico. U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-14/296, 2014.

Impact/Purpose:

The product describes a probability network modeled to estimate the effects on reservoir water storage capacity of two proposed management actions to protect coral reefs in southwestern Puerto Rico from sediment exposure.

Description:

A Bayesian belief network (BBN) was developed to characterize the effects of sediment accumulation on the water storage capacity of Lago Lucchetti (located in southwest Puerto Rico) and to forecast the life expectancy (usefulness) of the reservoir under different management scenarios. The conceptual approach included water and sediment delivery from two sources, from the Lucchetti watershed and from a tunnel linking Lago Lucchetti to three upstream reservoirs. Variables in the model included precipitation and erosion factors (soil type, landscape slope, and land use) applied to the Lucchetti watershed and to watersheds of the upstream reservoirs. The lack of available data for water and sediment flows in the watershed and through tunnels connecting the reservoirs led to several unique methods for network data acquisition. Status quo model runs demonstrated that sediment trapping has continuously declined in all four reservoirs since their construction and that every year a greater proportion of sediment is moving downstream through tunnels or spillways. The model compared favorably with incidental measured data in the region. Sensitivity analysis demonstrated that current sediment accumulation in Lago Lucchetti can be attributed in large part to sediment erosion from the Lucchetti watershed with only minor influence (~8%) from upstream reservoirs.

URLs/Downloads:

Record Details:

Document IDs: