Skip to content

Latest commit

 

History

History
executable file
·
49 lines (38 loc) · 3.73 KB

README.md

File metadata and controls

executable file
·
49 lines (38 loc) · 3.73 KB
Raw

HDS

A modified version (v2) of the Hysteretic Depressional Storage (HDS) model to simulate prairie fill and spill mechanism.

The original HDS (v1) was based on the equations listed in Ahmed et al. (2023). This version of HDS (v2) is revised based on the equations presented by Clark and Shook (2022), which is cleaner and more robust than the one proposed in HDS (v1).

Notes on the implementation

HDS should be implemented as a module just before the routing module in the model (i.e., after all runoff fluxes have been calculated). By doing so, the modeller ensures that the runoff that is being routed (reaches the river network) is the net runoff (after subtracting the stored water in the depressions).

HDS mainly needs the following information:

  • input fluxes/forcing: runoff, precipitation, ET (for pond/depression evaporation) in mm/timestep.
  • Subbasin/grid properties: depressions area (m2), depressions volume (m3), and total catchment area (m2)
  • Parameters (currently hard coded): p (shape of the slope profile [-]), b (shape of the fractional contributing area curve [-]), tau (time constant linear reservoir to simulate seepage losses from depressions [days-1]), rCoef (runoff coefficient [-]).

All these information are required at the sub-basin or grid scale. Please note that the two parameters (tau and rCoef) are conceptual, and it is better if not used when coupling HDS with a hydrological model. These parameters were used for comparison purposes against a synthetic test case from Clark and Shook (2022). So, it is better to set their values to zero so that they do not affect the results. If the model has a way to handle seepage losses from water bodies, then that can be passed into the model as input flux, but it would require some modifications in the function input argument.

For the rest of the parameters (p and b), please leave their value unchanged as these are the typical values for the Smith Creek Basin.

Please be aware that the following variables are state variables within the code, and it is crucial to ensure their proper maintenance from one time step to another for proper bookkeeping:

variable description
vMin ! minimum pond volume below which contributing area is zero [m3]
pondVol ! pond volume [m3]

Change log:

Changes are sorted from newest to oldest:

Sep 2024

  • add depCatchAreaFrac to the code and calculate the integrated basinConAreaFrac

Jul 2024

  • Add a condition to prevent updating vMin if the pond is completely full.
  • add other conditions to limit pondVol to depVol (when completely filled) and assign any excess input as outflow.

Oct 27, 2023

  • Calculate adjusted ET and infiltation for mass balance closure when pond is dry (losses > pondVol)

Oct 19, 2023

  • add constrains to avoid pondVol being greater than depVol under extremely wet conditions

Oct 18, 2023

  • add if statements to prevent negative pondVol values under extremely dry conditions

Sep 1, 2023

  • produce the outflow as an output timeseries.

Aug 30, 2023

  1. simplify code to combine runDepression and runOnestep into one subroutine runDepression and remove runOnestep.
  2. Produce the pond area as an output timeseries.
  3. Remove extra internal variables that are not needed.

Aug 30, 2023

  1. Modify the makefile to activate release and debug targets. The previous version only used release by default as the target and debug was not activated.
  2. Fix indentation (space -> tab) in ascii_util.f90 and type_HDS.f90.