Numbering Watersheds

Question

I am working with a watershed map, and I was wondering if there is a function or a way that I can get ArcMap to number the watersheds. For instance, if we have watershed X numbered as 0, I want the watersheds that surround/or share a boundary with watershed X to be numbered 1, and the watersheds that surround/ or share a boundary with the number 1 watersheds to be numbered 2 and so on. Is there a way this can be done by building a model in Modelbuilder? I would like to save it so when I change a watershed layer and run the model it will do the same in a new watershed area.

Thanks, E

PS: Sorry if the image is too small I did it really quickly using paint...

Answer 1

I'm thinking you could do something with Select by Location, Calculate Field and iteration in ModelBuilder... see this page for an example: Iteration using a count or Boolean condition

Answer 2

A possible solution written in arcpy Python based on @blah238's solution above. If you wanted to stay in a model builder environment, you could just take the expression from the CalculateFieldManagement line and use that in model builder. Just note the extra setting of -1 and 0 to the original field in case the input is a shapefile.

import arcpy
import sys

def watershed_counter(watershed_lyr, calc_attribute):
    current_count = int(arcpy.GetCount_management(watershed_lyr).getOutput(0))
    assert current_count == 1, "Too many features selected!"

    watershed_value = 0

    arcpy.AddMessage("Calculating base watershed")

    #We calculate to -1 as if base file is shapefile it uses 0 instead of Null
    #as placeholders for no data, and that'a where we want to start the
    #watershed count
    arcpy.CalculateField_management(watershed_lyr, calc_attribute, -1)

    while watershed_value >= 0:
        arcpy.AddMessage("Selecting new watersheds")
        arcpy.SelectLayerByLocation_management(watershed_lyr, 'SHARE_A_LINE_SEGMENT_WITH', watershed_lyr)

        new_count = int(arcpy.GetCount_management(watershed_lyr).getOutput(0))
        if current_count == new_count:
            watershed_value = -1
            arcpy.AddMessage("No more watersheds")
            break
        else:
            watershed_value += 1
            arcpy.AddMessage("At watershed level %i, %i new watersheds found" % (watershed_value, new_count - current_count))
            current_count = new_count

        data = {"val": watershed_value, "field": calc_attribute}
        arcpy.CalculateField_management(watershed_lyr, calc_attribute, "%(val)i if !%(field)s! is None or !%(field)s! == 0 else !%(field)s!" % data, "PYTHON_9.3")

    arcpy.AddMessage("Setting base watershed to 0")
    arcpy.SelectLayerByAttribute_management(watershed_lyr, "NEW_SELECTION", " %s = -1" % arcpy.AddFieldDelimiters(watershed_lyr, calc_attribute))
    arcpy.CalculateField_management(watershed_lyr, calc_attribute, 0)


if __name__ == '__main__':
    try:
        watershed_counter(*sys.argv[1:])
    except Exception, e:
        arcpy.AddError(e)
        sys.exit(e)

Note that each iteration of the loop in this solution is likely to take a longer amount of time as the size of select by location increases in complexity based on number of features, as well as the number of features to be calculated increasing.

Lastly if you're interested in having multiple possible starting points (ie, shortest destance), remove the assert in the script.