Values for slope length (SLSUBBSN) are discrete and variable across subbasins

[Steven Jepsen]

All,

In my SWAT model, built in ArcSWAT, HRU-values of slope length (parameter SLSUBBSN) take on discrete values. These discrete values vary across subbasins. Two things seem strange about this observation: (1) slope lengths are not supposed to vary within subbasins, at least according to the input/output manual, (2) the discrete nature of slope lengths suggests to me that there are rules at play in ArcSWAT for determining slope length, rules which may or may not depend on the size of subbasins. Maybe point #2 isn't all that strange considering that slope length likely depends on factors other than subbasin size. 

For example, slope lengths (m) take in these discrete values in HRUs of my model (N ~ 460 HRUs):

   9.14634

 60.97561

 24.39024

121.95122 (water)

This seems somewhat peculiar. Slope lengths therefore fall into classes. 

Does anyone know off-hand of the rule that is used to determine the classes into with slope lengths fall? 

I suspect it has to do with both land use and Hydrologic Soil Group. I'll keep examining the hru table of the database longer to see if I can find the pattern.

Steve

[Steven Jepsen]

All,

I think I found the answer to my own question from yesterday. In ArcSWAT 2012.10_2.15, assigned values of the HRU-parameter "slope length" (SLSUBBSN in *hru files) are dependent on the value of the parameter "slope steepness" (HRU_SLP in *.hru files). These SLSUBBSN-values take on discrete values of approximately 9, 24, 61, and 122 m, and are determined using a lookup table of HRU_SLP-values (Vigiak et al., 2015). Although I am not certain what the exact threshold values of HRU_SLP are (does anyone know?), I think they are close to 0.04, 0.12, and 0.20 (see plot below).

Vigiak, O., Malagó, A., Bouraoui, F., Vanmaercke, M., & Poesen, J. (2015). Adapting SWAT hillslope erosion model to predict sediment concentrations and yields in large Basins. Science of The Total Environment, 538, 855–875. https://doi.org/10.1016/j.scitotenv.2015.08.095

Steve

[Natalja C.]

Hi Steve,

It is good that you found the answer by yourself :) As I often state, the answer is usually found in the I/O or the theoretical documentation.

For the threshold values I often turn to Swat-cup Absolute_SWAT_Values file. There, the HRU_SLP is stated to be between 0 and 1. Not sure how physically accurate it is, but seems that these values are accepted by SWAT.

Best,

Natalja

[Steven Jepsen]

Natalia,

The three threshold values I am talking about, shown in my plot (i.e., HRU_SLP ~ 0.04, 0.12, 0.20), are not listed in the SWAT documentation (to my knowledge). This is where I first looked. These thresholds are in lookup table hidden somewhere inside ArcSWAT.

Steve 

[Natalja C.]

Well, the ArcSWAT documentation has some insights. Here is what it says on the slope definition:

during the HRU analysis there is an option to determine the amount of slope classes. Slope values are determined by the DEM.

So, if you set the slope classes to be , i.e. 4, and the maximum slope (derived from DEM) is i.e. 0.4, then your classes would be 0-0.1, 0.11-0.2, 0.210.3, 0.31-0.4, or something similar. I guess only one, who knows exactly how ArcSWAT calculates those things will know the answer. This is only a guess of mine :)

Best of luck!

Natalja

[Steven Jepsen]

All,

Finally, I was able to find  some specifics about the lookup table that SWAT 2012 uses by default to determine slope length as a function of slope sleepness. Part of this is based on Malagò et al. (2018), part on personal experimentation.

Here is the final lookup table:

slope steepness (-), slope length (m)

< 0.02, 122.0 m

0.02-0.12, 61.0 m

0.12-0.25, 24.4 m

>0.25, 9.1 m

Malagò, A., Vigiak, O., Bouraoui, F., Pagliero, L., & Franchini, M. (2018). The hillslope length impact on SWAT streamflow prediction in large basins. Journal of Environmental Informatics, 1–16. https://doi.org/10.3808/jei.201700384

Steve

On Tuesday, September 25, 2018 at 3:28:47 PM UTC-7, Steven Jepsen wrote:

[Jim Almendinger]

Steve --

Thanks for digging into this and posting the result -- very interesting and useful.

Cheers,

-- Jim

--
You received this message because you are subscribed to the Google Groups "SWAT-user" group.
To unsubscribe from this group and stop receiving emails from it, send an email to swatuser+u...@googlegroups.com.
To post to this group, send email to swat...@googlegroups.com.
Visit this group at https://groups.google.com/group/swatuser.
For more options, visit https://groups.google.com/d/optout.

--

James E. Almendinger, PhD

Director, St. Croix Watershed Research Station

Tel: 651-433-5953 x 11

Email: jalmen...@smm.org