Appropriate model time-steps

Owned by teamcity (Deactivated)
Last updated: 24 Oct 2016 by SysAdmin eWaterVersion comment
2 min read

Selection of the time-step for a catchment file is done when creating Meteorological Templates, and when choosing from the list of available meteorological templates to use (when you select New from the File menu).

The selection of an appropriate time-step for modelling is a compromise between accuracy and run time/computing power requirements. A smaller time-step will require longer computing time, and this may be an issue, particularly when running on slower machines or using Catchment files with many nodes (for more information on managing computing time requirements, refer to Running and Saving music Simulations).

An appropriate time-step will depend not only on the Source Node or overall Catchment size, but also on the size of Treatment Nodes within the catchment. For example, running daily time-step simulations on a 2m wide grass swale of 50m length, will produce unreliable results for both the routing of flows, and the behaviour of pollutants (because the variation in both will occur at time-steps of considerably less than one day).

Therefore, selection of time-steps should be based on the size of the smallest Source Nodes and Treatment Nodes within the overall Catchment, using a simple three-step process:

  1. Calculate or estimate the time of concentration (tc) of the smallest sub-catchment within the model;
  2. Calculate or estimate the shortest expected detention time of proposed treatment measures; and
  3. Select a time-step which is equal to or smaller than the smaller of 1 and 2.
Note: Wherever possible, it is recommended that a 6 minute time-step be used, as it will output  the most accurate results.

 In MUSIC, the statistics, graphing and export of data can be undertaken at any time-step equal to or larger than the time-step used to run the model.  For example, you can report at a daily time-step, even when the model was run at a 6-minute time-step.

Warm-up period

A default warm-up period is added in source nodes, that is designed to overcome some of the problems associated with short simulation periods. With warm-up enabled, the initial conditions will have less influence on the statistical results produced for longer simulation periods. This can be set up for source nodes only, using Settings » Preferences.