Negative values in accumulated variables (e.g. prlr, rsds)

Hi @lpalma

As we discussed yesterday in the dev meeting, I have added a filter after loading the data to replace negative values in accumulated variables with 0.

After doing this I also realized that for precipitation, if units are left in mm/s, the calibration step will set all values to zero. This has also been reported here: external/cstools#85 (comment 161986)

My question is, what's the best approach to handle this? Should I convert precipitation to mm/day before calibrating, and then convert it to mm/s again after the calibration is done? Or do you have a different suggestion?

Thank you,

Victòria