Stream Stats Equations

[philipbaileynar]

We want a more automated approach for defining streamflow for the hydrologic context project type (that feeds into BRAT). I have downloaded all the peak (2yr recurrence interval) and low flow equations from the National Stream Stats database:

2024_08_26_nss_depvars.csv

Tasks 1

Add a column to this CSV called python_equation and convert these "Excel" equations into Python syntax that will run with the eval command. This should be fairly straightforward. The exponent operation is different etc, but most of the syntax should be unchanged. Here's where we do this in hydro:

https://github.com/Riverscapes/riverscapes-tools/blob/a64481874e269a9e7e592a0da543c17e00847549/packages/hydro/hydro/utils/hydrology.py#L118

Task 2

This is hard... abort if it gets out of hand...

The "equations" listed refer to parameters, such as PRECIP for precipitation and FORESTPC for percent forested. Is it possible to strip out all the parameter names and then rank them by the decending frequency with which they occur? So I expect the result to be something like the following. The values to the right of the variable names are counts that they occur in the equations.

PRECIP 87
FORESTPC 78
ELEV500 67
....

Essentially I want to know the top 10 to 20 parameters that occur in these equations. The next step will be adding the calculation of these parameters to Riverscapes Context model.

FYI @jtgilbert

[philipbaileynar]

This old video where I explain the NSS database is just for my reference.

FYI related to #277

[r-k-g]

Full results for Task 2: most_common.txt

[r-k-g]

Results for Task 1: updated_equations.csv

There are a few problems that could prevent the new last column from running smoothly in a python eval().

• The last three lines contain the syntax e#^(...). While one may presume this is equivalent to exp(...), the syntax does not match any known format that I or chatGPT could find. As such, this has been left unchanged.
• One expression containing LogN(...) was replaced by math.log(...). These should be equivalent, but note that the math library must be imported.
• All equations containing function calls that use commas (round(), LogN()) are surrounded by double quotes. This may or may not cause a problem.
• Not 100% certain that the intended behavior of round() in the equations matches how it functions in python

[philipbaileynar]

Thank you!