temperature seasonality
anto.r
What is temperature seasonality? I read it is the difference in the
seasonal temperature, but how is it calculated in the BIOCLIM variable
creation? it says std dev*100, but std dev of what and is it very
different from the annual range?
thx
anto
Arturo
exist high variability in the area you are working, but, if it has low
values then doesn´t exist variability, and the climate for that
variable and that grid its more stable
Peter D. Wilson
Temperature seasonality is one of three ways that represent seasonal
variation in temperature in the Bioclim variable set. They are typically
very, very highly correlated (i.e. highly redundant) and, although they
statistically measure slightly different aspects of seasonal variation
in temperatures, I usually choose one of the three. The three measures are:
1. Temperature seasonality: It is SUPPOSED to be the coefficient of
variation of mean monthly temperatures (see official variable
definitions at the home of Bioclim software:
http://fennerschool.anu.edu.au/publications/software/anuclim/doc/bioclim.html#paramList),
but when developing the WorldClim data set a shortcut was introduced
(i.e. 100*Std deviation of mean monthly temperatures in Celcius) which
has dominated bioclimatic modelling ever since. The technically correct
way to calculate temperature seasonality is:
a. Convert temperatures to the absolute temperature scale by adding
273.15 (assuming degrees Celcius)
b. Calculate the coefficient of variation: 100*(std dev of absolute mean
monthly temperatures/mean of the mean monthly temperatures)
2. Isothermality: Annual mean temperature/Mean monthly temp range
3. Annual temperature range: Max temp warmest period - Min temp coldest
period.
Cheers,
Peter