Need exact locations for given flow accumulation values

4391
4
01-17-2011 03:44 PM
EgorNozdrya
New Contributor III
I am trying to generate sediment sampling locations as points for exact predefined flow accumulation values (i.e 2750 cells, 5500 cells etc). Those exact values are not in the flow accumulation grid - they have to be interpolated from neighboring values...

Can I achieve this with:
  (1) just flow accumulation grid
      or
  (2) should I generate measured stream network from flow accumulation grid and generate point events for given flow accumulation values (still not sure if it is possible)?

Any suggestions?
0 Kudos
4 Replies
EgorNozdrya
New Contributor III
select cells according to the focal max and focal min of a 3 x 3 neighborhood: cells where the focal max exceeds 2750 and focal min is less than 2750 are cells through which the 2750 contour line must pass, for example.


Nope, does not help!

I have tried and both things do not work - generating whole stream network below given value of 2777, not a desired points (see snapshot).
Contour lines are producing just mess around every pixel above desired value (2777 in my case)

Egor
0 Kudos
EgorNozdrya
New Contributor III
Nope, does not help!

Also tried to NULL flow accumulation grid outside of streams - does not help to get those points.
0 Kudos
MarkBoucher
Occasional Contributor III

The problem with the FAC values is that they seldom increase by 1. You will only find your exact FAC value if you are really really lucky.

Try this:

Use the Con() function to set a 1 for those FAC values higher than you target (lower in the watershed).

Use the Con() function to set a 1 for those FAC values lower than your target (higher in the watershed).

Convert the two resulting rasters to polygons. There will be a gap between the two polygons or their edges will touch. This is where your target would exist if it did exist. It would be the closets you will get to the answer.

0 Kudos
EgorNozdrya
New Contributor III

Thank you Mark!

That's really old topic - nearly 4 yers old)

Raster for polygons would take ages - I am working with huge NED DEM datasets.
Any raster to vector conversions are exceptionally slow for monster datasets.

As I remember - looking at old saved toolbox -  I've finally used to very simple solution suggested earlier:
(1) used Focal statistics with rectangle 2x2 cell filter to generate FocalMaxFlowAccumulation and FocalMinFlowAccumulation grids.
(2) calculated raster where  (FocalMaxFlowAccumulation >= ExactValue) AND (FocalMinFlowAccumulation < ExactValue)

(3) Converted cells of step2 raster to points - voila - have found exact locations for given drainage flow accumulation values.

Below is the snapshot of quick-and-dirty model to design stream sediment sampling program for 9 different survey resolution / initial drainage target size scenarios.

Clipboard01.png!

0 Kudos