An issue for @har917 @AlisonBennett (and maybe @juergenknauer since it touches on the CRU-TRENDY code).
The calculation of the precipitation in the met variable which is passed to our (bio)geophysics looks while in the presence of snow looks questionable to me. The precipitation is set to the sum of the precipitation and snow returned by the weather generator, see here. This appears to be different to the approach taken in the CRU, see here. CRU appears to circumvent any subdiurnal snow calculation by setting a blanket WG%SnowDay = 0.0, and then doing a post subdiurnalisation step to convert precipitation to snow where the temperature is below -2.
It may be that given BIOS is australia based, we hit these negative temperatures so rarely that it has little effect on the result, but we should make it clear precisely what these generated weather values actually are so they can be used correctly in the (bio)geophysics.
An issue for @har917 @AlisonBennett (and maybe @juergenknauer since it touches on the CRU-TRENDY code).
The calculation of the precipitation in the
metvariable which is passed to our (bio)geophysics looks while in the presence of snow looks questionable to me. The precipitation is set to the sum of the precipitation and snow returned by the weather generator, see here. This appears to be different to the approach taken in the CRU, see here. CRU appears to circumvent any subdiurnal snow calculation by setting a blanketWG%SnowDay = 0.0, and then doing a post subdiurnalisation step to convert precipitation to snow where the temperature is below -2.It may be that given BIOS is australia based, we hit these negative temperatures so rarely that it has little effect on the result, but we should make it clear precisely what these generated weather values actually are so they can be used correctly in the (bio)geophysics.