The PV dataset has a floating point issue. PV values change many orders of magnitude with height. Because I downloaded all levels at once and the ERA5 archive only has 32 bit precision, the small values in the lower troposphere have discrete values, which makes them kind of useless. If you want to use PV my advice would be to use the download instructions to download single levels at a time.
The PV dataset has a floating point issue. PV values change many orders of magnitude with height. Because I downloaded all levels at once and the ERA5 archive only has 32 bit precision, the small values in the lower troposphere have discrete values, which makes them kind of useless. If you want to use PV my advice would be to use the download instructions to download single levels at a time.