"John Mackenzie" wrote in message
...

snip
... the historic mean relative humidity figure I am
working from is 81%. I recorded a monthly mean for January of 80.42%.
The relevant summary page on my website does indeed show this as an
anomoly of -0.58 as well as 99.3% (of the LTA). I suppose what you're
saying is that having 2 % figures side by side representing different
things is confusing - and I can see what you mean. I 'll check into
what conventions there are in this area.

John ... the convention is that for an element which is aggregated
(i.e. has a monthly total) like rainfall or sunshine you express the
variation from normal as a percentage (either an absolute
percentage or difference from normal percentage ... I prefer
the former), while for an element which is meaned (i.e. as a
monthly average) like temperature or RH you express the
variation from normal as the difference between the mean for the
month in question and the long-term mean.

e.g. Mean temperature for Jan 2005: 6.1ºC (+1.9 degC)
Total rainfall for Jan 2005: 60mm (65%)
Total sunshine for Jan 2005: 65hr (118%)

My preference is also to use different units to express an
actual temperature (ºC) and a temperature difference (degC).
This is the Royal Met Soc house-style.

HTH .... Philip Eden