I am trying to develop a Bayesian Network model of water quality in an estuary using 10 years of monitoring data. The problem I am having is that the estuary has 3 different sources of water – each with different flow regimes and nutrient loading. I would like to have an early parent node with the percent contribution from each source but not sure how (other than using it as a scenario).
Absent guidance I will likely model each separately, flows or Q from each of the 3 sources (with the first bin being 0), nutrient loads TN and TP from each of the 3 sources. Then as child nodes pool flows as total freshwater input and nutrients as total loads. Then parse concentrations of TN, TP based on the loads and salinity, TOC, color, pH, TSS, turbidity all within one reach of the estuary; linking several of these to Chla, or optical model for estimating PAR. Having 3 separate sources of input is adding the complexity.
- Badges
- Science topic