All,

Attached is the illuminance (global horizontal) for Boise, ID, from
TMY2 data. Note the valley in the graph on the left bottom corner
which is caused by the very low illuminance values for January (the
values are in the hundreds, compared to tens of thousand in February).
I know for a fact that in Boise January is not (much) darker than
December or February.

Furthermore, the corresponding radiation data do not have that dip in
January. The illumination values are calculated based on the radiation
data, right? So if there is nothing wrong in the radiation data, is
this related to the model used in converting radiation data into
illumination values?

I also checked another site in Idaho, it has the same dark January.

Any thoughts?

Thanks,

Ery