Urban stormwater retention with snow


I’m using the Urban Stormwater Retention model for some canadian cities. I wonder if it is adequate to use the model considering that there are several months with snow. Rather than using the average annual precipitation, is it better to use the average annual precipitation of the rainy months only (spring, summer and fall)? Do we have to adjust certain input data to consider snow?

Thanks !


1 Like

Hi @anne-sophie.lessard,

Thanks for posting and welcome to the forums! Sorry this hasn’t garnered any attraction yet.

Maybe @Perrine , @b.janke , or @chris would be able to speak to this!



Hello Anne-Sophie,
Sorry for being so slow to reply. I have a few answers to your question:

  • First, the model was tested in Minneapolis-St. Paul, Minnesota (USA), where we have a few months of snowfall (in average years, anyway). The model performed pretty well using annual precipitation, with snow included as water equivalent. You can see this application in the end section of the User Guide for the Stormwater Model.
  • Second, in the included biophysical parameter table, the runoff coefficients are relevant to annual precipitation. They were determined using a SWMM model with a 10-year continuous simulation, including winter, using climate data from Minneapolis-St. Paul (2008-2017, if I remember right). This procedure is also outlined very briefly in the User Guide if you wanted to adapt it to your cities’ climates, but that section could admittedly use some refinement.
  • Something I had not thought of until you brought it up, however, is that the runoff concentrations are primarily relevant to the “warm” season, as that’s when most of those data were collected. For nitrogen and phosphorus, in my own work we have observed substantial seasonality in runoff concentrations (with some very high concentrations in urban snowmelt runoff in early spring / late winter), and this winter effect is not really included in the mean concentrations in the parameter table. Since runoff volumes tend to be lower in snowmelt, the effect overall might be small.

The short answer is that you don’t need to do anything to include the effect of snow, as it’s sort of included already if you’re using the default biophysical table. But it might be best to consider the model’s output as slightly conservative (slight under-prediction of retention and runoff loads) if you’re in a climate that gets appreciably more snow than Minneapolis.

Hope this is still timely enough to be helpful!
– Ben


Hi Ben!
Yes, your answer is still timely enough to be helpful! Thank you for these details and for your long and short answers, it is quite relevant.

1 Like