Views:

Question
I have precipitation timeseries data from multiple gauge stations in my study area. How can I implement them using the Rainfall Runoff model in MIKE+?

Answer

You can manually assign a rainfall timeseries to a selected catchment or you can let MIKE+ calculate the catchment rainfall for you based on precipitation timeseries from respective gauge stations which may be available in the wider area.

Follow the main steps to model rainfall-runoff when having multiple precipitation timeseries from gauges in catchment(s) in MIKE+, using e.g. the ‘RDI’ hydrological model:

Step #01 - Activation of "Catchment" option
At the ‘Model type’ editor (under ‘General settings’ at the ‘Setup’ data tree) activate the ‘Catchments’ option under ‘Features’ and the ‘Rainfall-Runoff (RR)’ option under ‘Modules’ – see Fig. 1.


Fig. 1 - The ‘Catchments’ Feature and ‘Rainfall-Runoff (RR)’ Module need to be activated in ‘Model type’ editor

 

Step #02 - Definition of "Catchments"
Define the ‘Catchments’ (and their areas) and apply e.g. the ‘RDI’ Hydrological model and the associated RDI parameters – see Fig. 2.


Fig. 2 - The ID, Hydological model, area and associated RDI parameters need to be defined for each catchment in ‘Catchments’ and ‘Parameters RDI’ editors

Step #03 - Specification of "Spacial extent"
At the ‘Boundary conditions’ editor, there are different options to specify the ‘
Spatial extent’ of the Rainfall timeseries:

  • All: You assign one rainfall timeseries for all Catchments
  • List: You assign one rainfall timeseries for the selected Catchments. You need to create a List of the selected Catchments first in Selection Manager
  • Individual: You assign a rainfall timeseries for each catchment.
  • Data source location:  You provide the coordinates of each station (Fig. 3) Here you don't assign the rainfall timeseries to a specific catchment.
  • Grid distributed weights: You apply a grid rainfall (dsf2).

As an example, in Fig. 3, a boundary for a rainfall station was inserted and the 'Data source location' option was applied in the 'Spatial extent' tab of the Boundary conditions editor. 


Fig. 3 - The ‘Spatial extent’ (and any user-defined ‘Distributed weights’) for applying the rainfall boundaries needs to be defined in the ‘Boundary conditions’ editor

Step #04 - Choosing the "Distribution type"
Choose the desired ‘Distribution type’ option in ‘Simulation setup’ > ‘Catchments’ tab - see Fig. 4:

  • The option ‘Closest rain gauge’, where only the closest precipitation gauge station is taken into account from the precipitation gauge coordinates defined using the 'Data source' option in the Boundary conditions editor.
  • The option ‘Thiessen polygons weighting’ (weights can be automatically computed or user-defined in the Boundary conditions editor > ‘Distributed weights’). For this option it is important that you have previously defined the ‘Data source’ location coordinates for each precipitation gauge station in the ‘Boundary conditions’ editor (‘Spatial extent’ tab).
  • The ‘Inverse distance weighting’ interpolation option (weights can be automatically computed or user-defined in the Boundary conditions editor > Distributed weights). Also for this option it is important that you have defined the ‘Data source’ location coordinates for each precipitation gauge station in the ‘Boundary conditions’ editor (‘Spatial extent’ tab).


Fig. 4 - In Simulation setup editor, ‘Catchments’ tab, the user needs to define the type of 'Spatial distribution of rainfall'

 

Important notes:

  • When importing a MIKE 11 setup into MIKE+, MIKE+ computes (on-the-fly) the weighted timeseries and they are used as input (one per catchment), without importing the original weights.
  • If the weights are user-defined or modified after computing them, then clicking the 'Compute' button will result in losing all the manual changes (in which case clicking the 'Undo' button should help recovering the original weights).
  • Computing the weights is optional. If not computed, the engine will compute them at run time (unless they have been defined by the user).
  • Computing the weights for any rainfall boundary condition will have the effect of computing the weights for all rainfall boundaries (since the weights for one rainfall boundary condition depends on the location of the other rain gauges).
  • It is expected that if the ‘Spatial extent’ of all rainfall boundaries is not set to the option ‘Data source location’ (e.g. set to 'Individual') then the ‘Distributed weights’ tab becomes deactivated. On the contrary, if even one rainfall boundary is using the ‘Data source location’ Spatial extent option, then the ‘Distributed weights’ tab/options become activated.
  • The 'Distribution type' at the 'Spatial distribution of rainfall' (interpolation) option (in Simulation Setup > Catchments tab) refers to a global setting for the entire simulation (as such it cannot be defined in the 'Catchments' or 'Boundary conditions' editors which only hold individual features/data.
  • If the 'Distribution type' at the 'Spatial distribution of rainfall' (interpolation) option (in Simulation Setup > Catchments tab) is set to 'Closest rain gauge', then the weights in 'Distributed weights' tab in Boundary conditions editor are ignored because they are not used for this method.
  • Only the 'Distribution type' option (in Simulation setup editor > Catchments tab) of the 'active simulation' is applied. So if there are multiple simulations defined in 'Simulation setup' editor, only the parameters of the 'active simulation' will be taken into account.
  • The option 'Grid distributed weights' in the 'Boundary conditions' editor requires the input of either timeseries (dfs0) or time-varying grid (dfs2) at the 'Temporal variation' tab. This option is designed to work with 2D time-varying grid files in dfs2 format, such as radar data.

Purpose and Conclusion
The purpose of applying precipitation gauge timeseries using the Rainfall-Runoff model is to better represent the spatial and temporal distribution of the precipitation variable within the catchment(s). Weights can be automatically computed or manually defined/adjusted by the user. The output will include weighted timeseries for the catchment(s).


FURTHER INFORMATION AND USEFUL LINKS

Manuals and User Guides
MIKE+ – River Network Modelling User Guide

Training options
MIKE+ – Getting started with river modelling

 

 
Related Products: MIKE+