When I checked the “swvl1” values in ERA5 for July 24, 2023 at 00:00 UTC, I was surprised to find 730 points below zero. Do these values have any special meaning?
The similar problem shows up in “swvl2” and “swvl3” too。
1 Like
I have been experiencing a similar bug where at least 40% of my data for the top 3 layers of soil moisture monthly values have larger than expected negative values. It appears to be a systemic issue with the data rather than a data processing issue on my end. Has anyone been able to fix this issue?
Since ERA5 is the product of model interpolation, I wonder if it’s appropriate to always force negative values to zero and assume that any negative moisture is simply “perfectly dry”. No model is perfect and I imagine soil moisture models are not finely tuned for ultra-dry conditions.