leaf-on/leaf-off seasonal variance

> It's right to turn on the groundwater module when the RUNOFF Block is adapted to a place with low DCIA. However, the existing RUNOFF does not simulate the interception process of canopies. The features that "antecedent conditions" including the "leaf off" (winter) / "leaf on" and seasonally varied interception storage suggested by Mr. Don Waye of NVPDC (12/15/98) cannot be accomplished by just tuning the coefficients in groundwater module.

> If the model cannot simulate a series of storm events and several continuous rain data sets, it is difficult to justify that the model is able to simulate the physical system. Particularly one time curve-fitted results can not guarantee the model calibration is realistic.

> Ching Lin Chen 12/28/98


Dr. Chen & other interested SWMMers,

There is much food for thought here. From a practical standpoint, I see 2 ways to go.

1. The most complicated, but maybe the best, way is to adapt the more robust IWMM/WARMF routines to SWMM. Dr. Chen & I may want SWMM usage confined to urban catchments, but its user base is quite large, and--like it or not--it has been applied beyond the urban realm. So, while many users already consider SWMM to be "a real watershed model", it could certainly be improved upon, especially in its handling of infiltration & groundwater flow (not to mention its water quality routines). I vote for improving SWMM.

Systech, CDM, and a handful of other firms, are well qualified to do this work, but of course nothing is likely to get done without adequate budgeting. EPA support seems to be in name only these days, but might there be opportunities via contract add-ons/task orders and grants??

2. The other way to go is to add a simple new input group in the RUNOFF Block that allows users to vary several parameters seasonally. This wouldn't get at all the concerns mentioned, but might go a long way. In other words, it seems useful to "keep it simple", as the greater question of the adequacy of SWMM's groundwater routines may be a separable--and complicating--issue.

I envision allowing seasonal adjustments to DCIA to account for the canopy effect during summer leaf-on conditions. If done through the "H" input group, the user could specify a different impervious range for each subcatchment. Or the user could provide a variance curve through specified monthly values and list the subcatchments the curve applies to (similar to the way evaporation is computed). A similar approach could be applied to pervious area depression storage to simulate leaf-on/leaf-off interception variance. Since overland flow often includes wooded/shrubby areas that vary seasonally, it may even be useful to consider varying the pervious area roughness coefficient, WW(6). I would go so far as to advocate varying EXTRAN/TRANSPORT channel roughness seasonally for many open channels, especially where they include vegetated, unmaintained floodplains in temperate regions.

By making these relatively simple adjustments to SWMM, I believe most of Dr. Ball's and my concerns can be addressed. It seems like perfect graduate sudent fodder to me. Any takers?

Don Waye,
NVPDC

Dear SWMMers,

Happy New Year to all. It was interesting to see my request for data on urban imperviousness evolve into a discussion of seasonal variation in effective imperviousness, although I was disappointed that no one offered any calibrated imperviousness data for a major city. I did try to use Figure 4-17 in the SWMM4 manual to evaluate my highly impervious urban watershed, but came out off the scale at 120% or so.

As for seasonal imperviousness, I have not noticed any such signal in any of my studies, but then I've never tried to calibrate RUNOFF in a non-urban setting, and rarely have year-long data. The one decidedly non-urban catchment I've modeled in the Boston area was very difficult to calibrate, but I attributed the problem to that (at least in part) I wasn't using the groundwater routines. The catchment had extensive wetlands, and the only way I could get the right shape to my long hydrographs was to lower the ground slope to a fraction of its apparent value.

That said, it would seem that the approach taken in TRANSPORT and EXTRAN for dry weather flows would meet the needs previously expressed by Don Waye, Ching Lin Chen, et al. In TRANSPORT, DWF is specified as a constant at each junction, but a series of multiplying factors can be applied. These are hourly, day of week, and monthly in TRANSPORT, but of course only monthly would be needed here. It's even possible to have different sets of factors for different parts of the system. These are the B4 lines in TRANSPORT and the BD lines in EXTRAN that (I think) Chuck Moore of CDM has coded. So if some one comes up with data to convince us all of its validity, I'll bet it could easily be added to RUNOFF (especially as RUNOFF only uses four B lines - plenty of room for expansion!)

One more thing - I recently retrieved two excellent discussions from the SWMM-USERS archives on I-D-F curve analysis and SWMM vs. MOUSE et al. These both made excellent references for information I needed, and were easily forwarded to colleagues around the world. Thanks to Rob and Bill James for making it possible, and to all of you for your insights.

Regards,
Mitch Heineman
heinemanm@cdm.com