Define raster extent

Question

I have a raster data set I imported to ArcGIS from a netCDF file. ArcGIS doesn't read the projection or extent correctly due to issues with the input file. However, I have the correct values for these parameters. Using the Define Projection tool, I am able to set the coordinate system without issue. However, the raster appears in the wrong place on the map (centered at (0,0)).

I am looking for an ArcPy tool that would allow me to define the extent of the raster manually in the same way that I can define the projection using Define Projection.

What I have tried:

• I can manually edit the extent in the properties window. However, I need to update the extent for 100s of these files, so need a scriptable/arcpy solution.
• I looked at the Set Raster Properties tool. Although the Get Raster Properties tool can access the extent values (Top, Left, Right, Bottom), it does not appear that Set Raster Properties allows the extent properties to be modified.
• The Shift tool alone won't work because the dimensions of the raster need to be adjusted in addition to location.

Edit:

I was using the wrong projection definition. When I use the right one, the map is displayed in the correct location. Among other clues, a curved Canadian border on a Mercator projection (ugg) is pretty obviously wrong, in retrospect.

Answer 1

This turned out to be a case of a projection issue (see my edits). As noted by Vince, this shouldn't be this hard. Nevertheless, I was interested to see whether I could use ArcPy to achieve the same results as editing the extent manually in the raster properties window. I was able to do so using the Shift and Rescale tools.

in_rast = "my_raster.tif"
out_rast_temp = "my_raster_temp.tif"
out_rast_extent = "my_raster_extent.tif"

top_correct = 6728275.583
left_correct =-14246503.87
right_correct = -7270751.753
bottom_correct = 2654659.006

x_dim_correct = right_correct - left_correct
y_dim_correct = top_correct - bottom_correct

top_wrong = float(arcpy.GetRasterProperties_management(in_rast, "TOP").getOutput(0))
left_wrong = float(arcpy.GetRasterProperties_management(in_rast, "LEFT").getOutput(0))
right_wrong = float(arcpy.GetRasterProperties_management(in_rast, "RIGHT").getOutput(0))
bottom_wrong = float(arcpy.GetRasterProperties_management(in_rast, "BOTTOM").getOutput(0))

x_dim_wrong = right_wrong - left_wrong
y_dim_wrong = top_wrong - bottom_wrong

x_shift = left_correct - left_wrong
y_shift = bottom_correct - bottom_wrong

x_scale = x_dim_correct/x_dim_wrong
y_scale = y_dim_correct/y_dim_wrong

arcpy.Shift_management(in_raster=in_rast, out_raster=out_rast_temp, x_value=x_shift, y_value=y_shift, in_snap_raster="")
arcpy.Rescale_management(in_raster=out_rast_temp, out_raster=out_rast_extent, x_scale=x_scale, y_scale=y_scale)

Result after running Shift:

Result after running Rescale:

The result has an extent that is identical to the input "correct" extent.

Again, this is probably the wrong way to go 90% of the time (vs. checking for upstream issues), but perhaps this can be useful in some situations.

Answer 2

Two things:

1. Thanks, nrp1000 for your solution. I had been trying to write a script to automate georeferencing of 100+ .tiff files and this workflow slotted in just about perfectly.

2. For anyone that comes along with a similar problem that they need to batch/loop process, you'll need to get extents, Rescale, get new extents, and then Shift.