Inundation Mapping

[Grass package]

>    Edwin Welles writes:

>    I am interested in finding out about
>    previous work in inundation mapping,
>    that is depicting forecast flood areas
>    areally with a GIS.

I haven't actually done any such mapping, but I did use a flooding
problem in class about two years ago, so I have given the problem
some thought. Suppose you have the DTM and a map of the streams for
an area, then there are really two problems to solve:
- water levels are not identical all over the area, so a rise in water
levels cannot be modeled by simply slicing the DTM at one particular
elevation;
- flooding is a dynamic phenomenon, with a rise in water levels first
occurring upstream, and any rise downstream being dependent on time
and the volume of water that the upstream part of a valley is able to
take (either in its stream channel or through flooding)

As a first approach you might disregard the dynamism, and create a
floodwaterlevel surface that is, say, 2 meters above normal, by running
a program like r.neighbours (max) iteratively on a map containing DTM
values for the streams. Then run r.mapcalc to find the areas where
this surface is higher than the DTM - these are flooded.

To add dynamism, you would need to establish a time interval, a rate of
water volume increase (if flooding is caused by a factor such as 
meltwater) or rainfall during that interval, and the speed by which 
floodwater go downstream. Then in each time step you will need to know the
local rise in water levels at a number of locations along the streams,
calculating this from water volume and speed, and model flooding for
that time step using a variant of the method suggested above. Then 
advance the clock, recalculate the rise of water levels, and recalcul;ate
flooding.

Phew! The more I think of it, the more complex it becomes. Obviously,
any rise in water levels downstream is not just dependent on the two
factors mentioned above, but also on the capacity of upstream areas
to take flooding. I suppose you could even model that in grass, but 
you would need a *lot* of processing power!

Cheers,

	Martijn

P. Martijn van Leusen -------------------------------------------------------+
Leverhulme Research Fellow, The Wroxeter Hinterland Project                  |
University of Birmingham Field Archaeology Unit                              |
Edgbaston, Birmingham B15 2TT  Phone +44 121 414 5513  Fax +44 121 414 5516  |
E-mail P.M.van-Leusen at bham.ac.uk                                             |
URL http://www.bham.ac.uk/BUFAU/Staff/vanleupm.html__________________________|