inconsistency delineating watershed (contributing area) from a single point (cell)

Discussion created by lrklein on Dec 11, 2010
Latest reply on Jan 2, 2011 by awhite42
I am trying to delineate the upslope contributing area from any cell in a 10m dem. I am using the pour point feature with the watershed tool. From the flow accumulation raster, I know for each cell, how many other cells contribute flow. Now, I want to know where all those cells are. I create a feature class of the point (cell) I am interested in know where all the cells are that contribute to its total flow accumulation number. I then use this point (feature class) as the input, along with flow direction, in the watershed tool. The problem is that sometimes I get the correct number of cells in the output, other times I don't. And, I can't figure out why. For instance, I have a cell that has a flow accumulation value of 398. I then use the watershed tool and find the area of these 398 cells that contribute. Great. Then, I choose another cell I am interested in - say, one that has a flow accumulation value of 561 - using the same procedure that correctly identified the 398 cells - will now only show me 15 cells of the 561 that are suppose to be contributing. Why is it not showing me all 561?
Thank you,