Confluences are top level objects which contain a fixed collection of slots which represents its important hydrologic processes. The slots are (click for an explanation of the standard slot display):

• Inflow: This slot collects all of the sources of water entering the reservoir. This includes virgin or unregulated inflows and releases from other structures upstream.

  If the inflow slot has a user defined series, that series will behave as an unregulated flow into the confluence.

  If the inflow slot is linked to one or more upstream nodes, the release from those nodes will be added to the user series. The total of these will constitute the inflow into the confluence.

• Outflow: The outflow slot routes the total inflow to any downstream nodes.

As the name implies, the most usual employment of a confluence is to gather flows from different upstream sources into a single stream and pass that downstream.

The confluence responds to all of the messages in the meta algorithm and to no others. During all steps of the MA except task, it simply passes the message down to each of its slots in turn. For example, the begin time step response of a confluence is simply to send a BTS message to each of its slots.

During the task step of the MA, the confluence performs all calculations needed to insure that its inflow and its outflow are equal. In doing this, the confluence may treat either inflow or outflow as the dependent variable depending on the policy being implemented. Thus, if the confluence is operating to satisfy a downstream demand, it passes the request upstream. It is free to communicate with upstream water sources if necessary.