tickling r.watershed

[Keith Mitchell]

We are using runoff maps for input into
r.watershed so we can identify pathways 
and areas of runoff concentration.

that is : accumulation maps.

We are using a map layer that represents
mass of pesticide concentration in the runoff
as input.

Fine so far - well, we get mostly (you guessed)
negative numbers in the output, which means very
little to me.

Is  there any alternative to waiting hours for this
mumbo jumbo?  I've changed the basin values(t parameter
below, I installed a mask around the watershed, I even 
set the area byond our DEM so that 6the entire map 
is bound by zero values.

Perhaps someone knows how to use r.drain in a script?
I don't think r.mapcalc will be useful for this.

Also, the program runs ok without an input 
overland flow map - when you give it a runoff
layer it generates the stange output.

the command line we used was:
/data/grass/etc/water/ram el="e.lscaled2 at keith" ov="ms at cyananzine" t=100 ba="ms.basin" ar="ms.basi.file" ac="ms.acc" di="ms.acc.dis"

Keith M. Mitchell
University of Illinois
Department of Agronomy
S-316 Turner Hall
1102 S. Goodwin Ave.
Urbana, Il. 61801
(217) 333-9648
keith at piked2.agn.uiuc.edu