All Forums

sensitivity to slope

In running some basic long-term continuous simulations I’ve noticed that SWMM output may not be very sensitive to the slope subcatchment parameter.  For instance, I ran two simple continuous SWMM5 models with over 50-years of hourly rainfall data and a 1-acre catchment.  One model had a slope of 0.5% and the other 50%.  All other subcatchment parameters were the same.  While I would have thought that the 50% sloped catchment would produce more runoff than the 0.5% catchment, the results turned out to be about the same.  The runoff coefficient for 0.5% slope was 0.212 while that for 50% slope was 0.223, only a 5% increase.  The peak flow rate for 0.5% slope was 0.122 cfs while that for 50% was 0.124 cfs. 

 

My questions are:

  1. Do you think this slight increase in runoff volume is a realistic result or do you agree with me that it seems low?
  2. Are there more appropriate ways to set up a SWMM model to account for increased runoff caused by steeper slopes?
  3. Are there other continuous hydrologic models that are better at accounting for changes in runoff with slope (HSPF, HMS, etc)?
  4. Is the low sensitivity to slope due to any other parameters you can think of (time step, catchment area, etc)?

Thanks

You need to be a member of SWMM 5 or SWMM or EPASWMM - (not formally associated with EPA) to add comments!

Join SWMM 5 or SWMM or EPASWMM - (not formally associated with EPA)

Email me when people reply –

Replies

  • Rob, thanks for the reply.

     

    Separately, my attempt at

    something to check swmm5 input data quickly using firefox

    www.hydrocalc.com/swmm5_statistics/swmm5_statistics.html

     

  • The depression storage does not change as the slope changes

     

    in the SWMM 5 engine the flow from a subarea is (since all units internally are cfs)

     

    Runoff = 1.486 * Width /  N * Slope^0.5 * ( Depth - Depression Storage) ^ 1.666667

     

    for the three subareas:  Impervious, Pervious and Impervious without Depression Storage

  • To the OP:

    Please see the link:

    http://www.miduss.com/theory-Calculating-Runoff-SWMM-RUNOFF-Algorit...

    Urban Stormwater Modelling.. by S.J. Nix has detail of SWMM subcatchemt model.

     

    As I understand it,

    SWMM5 subcatchment runoff model is a rectangular trough. Increasing slope does not proportionally increase peak runoff rate. And, for totally impervious subcatchment, the slope is irrelevant to the volumetric runoff coefficient. (I am distinguising runoff flow rate and volumetric runoff coefficient.)

     

    My answers to the original questions:

    1. It is as realistic as the subcatchment model is modelled.

    2. The depression storage affects the volumetric runoff.

    3. I think SWMM subcatchment model is more physical based and mathematically correct. It is not an empirical model. Neither it is THE reality-it still is a model. But I am not that familiar with HSPF and HMS.

    4. Low sensitivity is because: steeper slope does not increase or decrease amount of water retained in the depression storage-at least in the way it has been modelled. I am slightly unsure of this. Rob, does the depression storage take into consideration the slope?

     

    (everything is from my understanding or misunderstanding of the matter. so don't take them seriously.)

  • You can also get some information about the sensitivity by looking at the runoff and rainfall together - if the runoff has the shape of the rainfall then the individual sensitivity of any parameter will be low.

    1058529800.png?width=721
  • Hi Read,

     

    The square root of the slope is used internally in SWMM 5 in the calculation of the  runoff flow.  The coefficient is called Wcon internally and 

     

    Wcon = 1.486/n * Width * Slope^1/2

     

    so a slope of 50 percent has only a 10 times faster runoff than a slope of 0.5 percent.  You did not mention the percentage of impervious but if most of the pervious rainfall ends up as infiltration then this would also have an impact on your total runoff.  The Wcon parameter also had more than one variable composing it - it your Width is large then the runoff may be already fast and changing the slope merely changes it from fast to faster.

     

    1058529805.png

This reply was deleted.