Valley-channel width ratio application to binary IP model outputs

[smnorris]

I am not sure if/how we can apply this measure to the existing binary spawning/rearing models. For highly confined Coho streams Burnett 2007 only bumps the IP model down to .2, it does not exclude the streams entirely.

• are there ratio thresholds at which we can presume spawning/rearing do not occur (or are minimal enough to exclude)?
• if not, do we need to modify the data model to support relative percentage scores for spawning/rearing potential?

I'll have to generate some valley-width index numbers and compare to known rearing/spawning locations as a basis for discussion.

[nickw-CWF]

The other parameters (gradient and MAD) are also continuous variables in traditional IP models. For setting existing thresholds, I compared scientific literature values for spawning/rearing with the high/medium value ranges from existing IP models, while excluding "low" ranges for each variable, to turn it into a binary measure of habitat potential. Part of this was to exclude low value habitat from consideration, but it was also to support quantification of the habitat connectivity status for a watershed. It is more involved to calculate a % connected if each stream segment has a suitability rating between 0-1, rather than true/false.

We could try and set min./max. thresholds on confinement values to maintain the model structure. See Sheer et al. 2009 for how channel confinement is used in other IP models.

Another option would be to also convert gradient and discharge to suitability scores (0-1), so that each stream segment gets assigned an overall habitat suitability score and then we set a cut off on for what we want to consider "suitable" habitat for quantification purposes (i.e., an overall score > 0.65 = potential suitable habitat).

I can talk to Nick L. about this, but it might be a good topic to bring to the group next week. Let me know if you'd like to hop on a quick call to discuss further.

[smnorris]

Thanks. I would lean towards option 1 for the short term but wondered if it is possible to define a valley-width index that rules out habitat potential. We will have to look at the data. That is a great reference for summarizing approaches.