“Valid percent” indicates the percent of pixels that are not nodata. Comparing “valid percent” values across rasters may help you identify cases of unexpected nodata.
Hello @Abhi0710 -
The “Valid percent” values for rasters includes all of the NoData pixels that are outside of your study area (but are needed to fill in the rectangle shape of a raster). So it is very common for this value to be smaller than expected. The main purpose for “Valid percent” is to compare between the input rasters, if one layer is much different than the others, it would be good to look into why, since that might mean that one layer does not cover the entire intended area of interest that you want to model.
Honestly, i also find this confusing, and i’ve asked for clarification to be added to the documentation for this field.
The best way to evaluate missing data within the study area is to look at a map with the NoData values symbolized and highlighted. This shows whether there are any NoData pixels within the study area, which will lead to NoData values in the modeling results, and is really what we care about.
~ Stacie
