This is less a problem than an inquiry into a better way of doing something. I am curious how people might go about selecting, say, all areas less than or equal to 10 feet (from a DEM) that are contiguous to the coastline. The purpose of this exercise would be to map low-lying areas along the coast that may flood due to surge during a coastal storm.
Here is an abridged version of how I've been doing this sort of thing:
1) Extract from DEM all values less than or equal to 10 feet using the (raster) clip tool.
2) Convert the raster to a polygon using the Raster to Polygon tool.
3) Select areas contiguous to the coastline layer:
a) Select all polygon areas that intersect the coast line. Export selection.
b) Add to selection all areas that intersect the new selection. Export.
c) Repeat this process until the difference in the number of attributes from one exported selection
to the next reaches 0, meaning, you are no longer selecting new features when you select
intersecting features.
4) Delete the remaining "orphaned" areas.
It's really step "3" that I'm trying to find an alternate method to. It can take up to 20 iterations of this step to select these areas for NYC. It just seems so tedious. I'm really curious how others might go about performing this task.
Update: I've come up with another method that eliminates the very iterative Step 3 from above. I am still curious to read how others might accomplish this task though.
Alternative Method:
1) Extract from DEM all values less than or equal to 10 feet using the (raster) clip tool.
2) Give every raster cell the same value (i.e. create a new field and have all values in that field = 1).
3) Convert raster to polygon using the Raster to Polygon tool.
4) Select and export features that intersect the coastline.
The above will get you your contiguous "flooded" area, but it will not have elevation values assigned to it, like the first step of steps above. To get those back, you'll need to do the following:
A) Use the resulting features (shapefile or feature class) as a mask to clip the raster.
B) Convert raster to polygon.
Best Answer
Apply Region Group in spatial analyst with this options:
Records that have match is what you need