r.watershed

[Bahram Saghafian]

In an E-mail message, Dan Riker wrote:
> 
> Hello, I hope all is well
> 
> I am presently working on TOPMODEL, a rainfall-runoff model, for my thesis.  What I need GRASS for is a topographic index ln(a/tanB), but I'm not certain how to calculate it.  I would also like to know  how to get the mean, std deviation and skewness of t> hat index. Finally I need to put it in an I*2 form to be used by the model.  
> If anyone has any information on any aspect of this I would certainly appreciate it.
> Dan Riker
> Duke University
> 

If I can recall correctly, parameter "a" is the specific catchment area
(the upslope area per unit contour width) and "B" is the bed slope angle.
To get "B" simply run r.slope.aspect. Computation of "a" isn't so simple.
For more reliable outcome either remove the pits from your region or 
collect them in a file to be used in the next step. Then run r.watershed
to get the accumulation file. Dividing accumulation by cell size closely
approximates "a". Disregard the negative values in accumulation file; they 
indicate that those cells actually drain out of your region. Having slope 
and specific area files use r.mapcalc to compute ln(a/tanB). 

I don't remember what program calculates mean and std dev. Nevertheless 
you can make an ascii version of your final output file by r.out.ascii 
and write your own code to do the job.

B.S.