Need help with calculating area for land cover

[NitishaShrestha]

Hi,

I am trying to calculate the area for each classification in NLCD land cover data.

I tried adding field in the attribute table but it does not let me use calculate geometry. I also tried using tabulate area from spatial analyst tool but that does not work either. Please help.

Thank You,
Nitisha

[NitishaShrestha]

For some reason, it didnt work earlier but it worked now.

When I exported it as a grid the new count is still 2630. Since the ohio dataset was clipped according to the buffer of the watershed  I think the attribute table of entire Ohio gets copied. So the new clipped dataset  shows the actual count from the Ohio not for my watershed. So now when I converted it to grid it shows the count for my watershed. I am not sure if I am making sense or this even happened ?

:confused:

[NitishaShrestha]

I have attached the attribute table for the original and clipped dataset

[MatthewNordhagen]

Hi Nitishia - Sometimes I'll have that same issue with Arc giving me the "failed to save raster data set" error.  More often than not it's usually something about the name of the raster that Arc doesn't like.  I'll end up trying a couple different variations of the name until it takes.

So, based on your attached attribute tables did you mask your data set instead of clipping the data set?  If you mask a raster the attribute table does not reflect those changes.  Instead, Masking defines all of the pixels outside of your watershed boundary as NoData.  The data is actually still there you just don't see it.  If in fact you did use the Clip tool under the Raster Toolset in Data Management then you will need to rebuild your attribute table to reflect any changes made to the new clipped raster. Arc will often carry over the original attribute table even though you've clipped the data set.  I should have also mentioned that when you create a raster with the Con tool, the tool will actually update all of the values in the new raster's attribute table to reflect only those pixels that were identified as open water in your SQL statement.  In the end your 2630 pixel count is correct given that you are only analyzing just a small watershed.  The very large pixel count that you see is for the entire state of Ohio.

Let me know if these clears things up for you and whether or not you need any more help!

[NitishaShrestha]

Hi Matthew,
Thank you for your help and time.

It finally worked. I had to check Use Renderer in the export raster data.

Once again thank you so much,
Nitisha

[NitishaShrestha]

Hi Matthew,

Changing the imagine file to grid and then to polygon worked for all the data sets except for NLCD 2001/2006 From �?? To Change Index. I have attached the attribute table for the data set (original_attributetable and grid_attribute table).

I have highlighted 'Value' 15 because it has 0 count in the original but in the grid its count is 27947. In other land cover, the value of grid represents the class code. So it was easier to re classify ( I have attached 2001 original and grid attribute table). Since the 'Value' in  2001/2006 From �?? To Change Index data set does not represent the class code i am not sure how to reclassify it (the class code for the original is in '2001_2006' column of original_attributetable). Any suggestions would be helpful.

Thank you so much for your help.
Nitisha