[GRASS-user] multiflow direction models

[Helena Mitasova]

Jus a short explanation - people often mix 2 things - flow direction  
and flow accumulation.

You can have the following flow/gradient direction:
D-4 (direction is discretized with 90 deg step), D-8 (discretized  
with 45 degree step), D-inf (discretization is
defined by FP precision)

The flow accumulation can be computed as
SFD (single flow direction - all water flow into single lowest cell),
MFD (multiple flow direction - water is distributed into multiple  
lower cells - different rules can be set),
  full 2D (bivariate routing according to continuity equation).

You can combine these in different ways, none of the methods in GRASS  
is exactly the same as Tarboton's.

Helena



On May 15, 2006, at 5:15 PM, Dylan Beaudette wrote:

> Kessie,
>
> Ah. Well this might be a good time to ask the GRASS developers to  
> clarify. You
> might also check out the docs behind r.terraflow .
>
> cheers,
>
> Dylan
>
> On Monday 15 May 2006 13:51, Ksenia Konwicki wrote:
>> Thanks Dylan.
>>
>> I followed this string fairly closely when it was active. I have
>> investigated r.flow and r.flowmd.
>>
>> Question though... I am a little confused by the r.flow  
>> documentation. It
>> states that it used the D-inf logarithm to calculate flow direction,
>> however flow accumulation uses a single flow direction logarithm. The
>> r.flowmd module appears also offer D-inf flow direction however it  
>> seems to
>> only offer the Desmet & Govers multiflow contributing area  
>> calculations for
>> flow
>> accumulation.
>>
>> Our research favours Tarboton's algorithms. We would like to use  
>> GRASS for
>> the computations however we will have to find a programmer or  
>> another Open
>> Source option if it is not available.
>>
>> Thank you,
>>
>> Kessie.
>>
>> Dylan Beaudette wrote:
>>> Kessie,
>>>
>>> Might check on some past GRASS-user messages on this thread:
>>>
>>> http://grass.itc.it/pipermail/grassuser/2006-February/032744.html
>>>
>>> Perhaps it would be a good time to create a page on this...
>>>
>>> Dylan
>>>
>>> On Monday 15 May 2006 11:34, Ksenia Konwicki wrote:
>>>> Good morning,
>>>>
>>>> I am wondering if anyone knows whether David Tarboton's D-infinity
>>>> multiflow direction algorithms have been integrated into a GRASS
>>>> module.
>>>>
>>>> Thank you,
>>>>
>>>> Kessie.
>>>>
>>>> --
>>>> Ksenia E. Konwicki, RPF
>>>> Ecologist
>>>>
>>>> Timberline Forest Inventory Consultants
>>>> 1579 9th Avenue
>>>> Prince George, BC   V2L 3R8
>>>> P (250)562-2628
>>>> F (250)562-6942
>>>> E kes at timberline.ca
>>>>
>>>>
>>>> _______________________________________________
>>>> grassuser mailing list
>>>> grassuser at grass.itc.it
>>>> http://grass.itc.it/mailman/listinfo/grassuser
>>>
>>> --
>>> Dylan Beaudette
>>> Soils and Biogeochemistry Graduate Group
>>> University of California at Davis
>>> 530.754.7341
>>
>> --
>> Ksenia E. Konwicki, RPF
>> Ecologist
>>
>> Timberline Forest Inventory Consultants
>> 1579 9th Avenue
>> Prince George, BC   V2L 3R8
>> P (250)562-2628
>> F (250)562-6942
>> E kes at timberline.ca
>
> -- 
> Dylan Beaudette
> Soils and Biogeochemistry Graduate Group
> University of California at Davis
> 530.754.7341