Uncertainty in these trend values arises from
different methodological choices made
by the
groups deriving satellite products (Mears et al., 2011) and radiosonde compilations (Thorne et al., 2011), and from fitting a linear trend to a time series containing substantial interannual and decadal variability (Santer et al., 2008; McKitrick et al., 2010).
This «about 3C» sensitivity is constrained
by a bunch of data and has been independently
derived by a bunch of
different groups employing a bunch of
different methods including a bunch that don't use GCMs at all.