Artificial Neural Networks (ANN)

In 1995 the Modeling Support Branch started using Artificial Neural Networks (ANNs) to quickly simulate the flow-salinity relationships in the Delta as an alternative to using DSM2. The ANNs are typically trained on the input and output of DSM2 simulations.

Improving the ANNs is an on-going process. As improvements are made within DSM2 or the estimation of its boundary conditions, new ANNs are trained. Currently the DSM2 ANN is used to estimate salinity within CALSIM. Other networks have been trained to represent other water quality relationships in the Delta (such as the THM simulator), but the most common use of ANNs by the branch is to estimate salinity compliance based on flow conditions within CALSIM.


Selected chapters from the annual Methodology for Flow and Salinity Estimates in the Sacramento-San Joaquin Delta and Suisun Marsh Reports
Other Reports:

Artificial Neural Networks with Application to the Sacramento-San Joaquin Delta. (March, 1995). California Department of Water Resources. Sacramento, CA.

Back | Home Computer Models Reports

Section Home Pages

Hydrology &
Delta Modeling Computer Assistance


Last modified: March 2, 2009.

California Department of Water Resources.
All rights reserved.