r.watershed modeling

[Luca Palmeri]

Hi everybody,

first of all I have to say that I am not an hydrologist so please excuse me for
eventually making errors or posing trivial questions.
It is a while now that I am puzling with a problem, and I cannot find a
solution so I ask to the list if there is somedbody who is actualy doing some
kind of hydrological watershed modeling using grass, and particular
r.watershed.
I am studying an hydrographical basin of environ 50000 ha. I have an enough
detailed description, with most maps on a 50 mt resolution. I have of course a
DEM at 50mt res. For my analysis I need an estimation of the *annual* discharge
in each cell of the basin. I have average-year precipitation data, a vector map
of channels and rivers (but without any data on the importance of  the water
flow but only 4 classes of rivers). I also have evapo-transpiration
estimations. I have no information about infilitration rates or hydraulic
resistance, apart maybe from the soil type that can be of some help in deducing
that.
The situation in the area I am studying is further complicated by the presence
of many drainage support systems for agriculture that alter substantially the
watershed flow. For these I have the location and the consumption in kwh/year.

I am wondering how is it possible to link all these elements in order to obtain
at least some very rough estimation of the water that traverses each cell every
year.

I have been using r.watershed, reducing the resolution to 150 mt and using an
overland flow map that comes from the river map assuming for the different
types of rivers arbitrary values of the flow (in million cubic mt/year)
combined with rainfall minus the evapotranspiration. But the accumulation map I
get from r.watershed is for the most part empty, and where it is not  empty
contains mostly negative values. Also the basin map produced  by r.watershed is
quite far from the "real" sub-basins that I have from previous studies at
larger scale.
I have to say that I do not have information outside the are that I am studing,
or in some cases only very rough one.

I really do not think I have enough detail and time to use some tool like
r.hydro.CASC2D, and if I understand well r.water.fea wont be much of help. For
these reasons I decided to try it by using r.watershed.

If some body here uses r.watershed and/or has some suggestion on this subject,
any help will be greatly apreciated

thanks in advance

Luca

--
--> Luca Palmeri - Dip. Proc. Chimici dell'Ingengeria
    Universita' di Padova  TEL: +390498275527 -------
-------------------------  FAX: +390498275528 -------
--> Think holistic @ the end of the 2nd millenium <--