# High raster values with 100 Square kilometre area

392
4
10-13-2021 08:17 AM
by
New Contributor III

Hello all,

I have a raster with suitability value ranging from 0 till 1. 0 indicates least suitable and 1 is the best. In other words as value increases from 0 to 1, suitability also increases.  I am looking for an exactly 100 square kilometres of areas from this raster image with high suitability value(SV).

I am doing trial and error by reclassifying the raster image with 0.8 and above values and then checking the overall size from the attribute table.

If it is more than 100 Sq km then i make a another (higher) cut off in reclassifying like 0.85 and then checking the cell numbers of the result. Is there some other simpler way to do this than doing trial and error?

Thank you !

1 Solution

Accepted Solutions
MVP Frequent Contributor

If I was doing it very quickly I might look at the total extent (square km) of the raster and set a symbology type of Quantile, with the number of quantiles set against the % of the total extent which makes up 100 sq km.

Another way would be to look at or build a raster attribute table Build Raster Attribute Table (Data Management)—ArcGIS Pro | Documentation but you would first have to modify your data to not be Float.  Subsequently you would then multiply the cell size by the count, which would give you the total area corresponding to each cell value.  Just keep adding these up until you get to the break point of 100 sq km.

4 Replies
MVP Frequent Contributor

If I was doing it very quickly I might look at the total extent (square km) of the raster and set a symbology type of Quantile, with the number of quantiles set against the % of the total extent which makes up 100 sq km.

Another way would be to look at or build a raster attribute table Build Raster Attribute Table (Data Management)—ArcGIS Pro | Documentation but you would first have to modify your data to not be Float.  Subsequently you would then multiply the cell size by the count, which would give you the total area corresponding to each cell value.  Just keep adding these up until you get to the break point of 100 sq km.

MVP Esteemed Contributor

Unlikely unless you know the shape of the suitability - area curve.

... sort of retired...
by
New Contributor III

Thanks for all the inputs 🙂

Esri Contributor

Alternative solutions:

You can use Geoprocessing Locate Regions tool to identify the optimal contiguous location for the size you want that meets the specified shape/utility trade off. If you do not allow islands in the resulting area, you may or may not get the exact amount of area you want (it maybe a little more). See the tool help and how it works to get an understanding of the tool.

If you do not want the resulting locations to be contiguous and you just want a specified amount of area with the highest suitability, you can use the second approach identified by David. One way to do this is multiply the suitability raster by 100 or 1,000 and then apply the Int Spatial Analyst tool. An attribute table will be created with the resulting raster. One of the fields in the table will be Count (the number of cells with that values). Determine the area of a cell and identify how many cells you need to meet your area needs. Add the count of the highest value to the second highest value and so forth until you get the desired area, note the “Value” where this happens. Run the Extract by Attribute tool and specify the clause where “Value > …”. Depending on your analysis, you may get some suitability values with too many cells and when adding it next to the running total it may exceed the desired area. You can add more criteria to stratify the category.