An unsigned integer is always non-negative numbers.
A signed integer can store negative values.
But how are they used for raster data?
Unsigned Integer Raster
With an 8-bit unsigned raster, valid values are from 0 to 255. This means that an 8-bit raster has 256 values in total.
When you talk about valid ranges (or pixel depth), this is the radiometric resolution of a raster image.
Radiometric resolution is the amount of detail in each pixel expressed in units of bits. A wide range of values gives the ability for pixel values to discriminate very slight differences in energy.
All things being equal, an 8-bit raster produces significantly greater detail than a 4-bit raster.
The valid range for a 8-bit signed is -128 to 127. If you have values larger than 127 or less than -128 in the input, than you will need to use 16-bit signed.
Some raster formats do not support signed images:
Signed Integer vs Unsigned Raster Formats
When you convert an image in ArcGIS ( Data Management Tools > Raster > Raster Dataset > Copy Raster or another program, specify the pixel depth and file type.
When exporting as a signed raster, make note of the formats that support it.
- JPEG, JP2, BMP, GIF and PNG do not support 8-bit signed, 16-bit signed or beyond. This means that you won’t be able to store negative values with these raster formats.
- IMG, TIFF (GeoTIFF), GRID, JPEG, JP2, BMP, GIF, PNG, BIL/BIP/BSQ and DAT raster formats can store negative values.
- IMG, TIFF an GRID are the most versatile formats. These formats accept 8-bit unsigned, 8-bit signed, 16-bit unsigned, 16-bit signed, 32-bit signed and 32-bit floating.
Next time you’re selecting a raster output, carefully select whether or not it’s an unsigned integer or signed integer.
Unsigned numbers are always non-negative numbers.
Signed numbers can store negative values.
IMG, TIFF and GRID are acceptable formats for 8-bit, 16-bit and 32-bit signed integers.