I have a netcdf file (LOCA2 data) containing temperature values stored in the variable tasmax_tavg. Using the "Make Multidimensional Raster" in ArcGIS Pro 2.9.9 I get one set of stats, when using "Copy Raster" to crf I get different stat values. When I do the same process in 3.4.0 I get a whole new set of different values. When I run python xarray to get stats, I get values similar to 2.9.9 Copy Raster but not the same.
My environmental variables are WGS84, no resampling, no compression, no pyramids, calculate statistics.
I'm not sure what to try next. The balance of the analysis relies upon these transformations being consistent.
I'm willing to share data if someone wishes to corroborate my findings.
Below are the stat values for each of the processes:
| ArcGISPro 2.9.9 | Min | Max | Mean | StDev |
| Copy Raster | 257.16729736 | 324.49603271 | 294.33952381 | 10.98436202 |
| Make Multi | 272.35498047 | 306.82998657 | 290.69200234 | 6.78096632 |
| | | | | |
| ArcGIS Pro 3.4.0 | | | | |
| Copy Raster | 271.18478394 | 312.28817749 | 293.33416138 | 6.88249549 |
| Make Multi | 272.35498047 | 306.82998657 | 290.69200234 | 6.78096632 |
| | | | | |
| Python xarray | 257.16729736 | 324.49603271 | 294.33895874 | 10.98434544 |
