deletions | additions       diff --git a/subsection_textbf_Comparison_Not_really__.tex b/subsection_textbf_Comparison_Not_really__.tex  index e91832a..d540a3a 100644  --- a/subsection_textbf_Comparison_Not_really__.tex  +++ b/subsection_textbf_Comparison_Not_really__.tex 
...
\\  Not really sure how to best do this, so below is basically a whole bunch of comparisons which I'll then try and decide which bests shows what we want to know - how much does the bias in the CFSv2 H_2O column influenced the modeled $\delta$^2H_{column}?  Figures \ref{fig:MeanProfiles} and \ref{fig:MeanProfileBias} show the mean (\pm1$\sigma$) profiles for each season and the bias (CFSv2 - Sonde), respectively. \\  \textbf{DJF}: water vapour profiles show a narrow distribution throughout the troposphere with an almost linear decrease in mixing ratio with height. The 2 datasets agree within ~1mmol, with the largest biases for the CFSv2 dataset at the surface (dry bias) and between 800 and 400mb. However, the middle troposphere bias decreases when the RH corrections are applied to the sonde data. \\  \textbf{MAM and SON}: The transitional seasons show a mostly linear decrease with height for the mixing ratio, however, there is more variability. MAM and SON have wet biases in the lower free atmosphere (800-600mb), and a dry bias for SON at the surface. The lower free atmosphere bias decreases with corrections applied to the Sonde data.\\  \textbf{JJA}: The dry season shows a clear boundary layer, with a steep gradient in the mixing ratio around 800mb. Probably the result of the descending branch of the Hadley cell being over Darwin during this period. The bias is greatest at the surface (dry) and lower free atmosphere (wet). Here the RH corrections have little impact on the bias between the 2 datasets.  The RH shows similar features in the bias relative to the mixing ratio. However, these is a large wet bias in the UTLS region.  \citet{Wang_2010} show a positive bias in relative humidity for RH  in the UTLS for several  reanalysis datasets  relative to sondes (drift sondes in their case).This is similar to what we observe for Darwin (see figure \ref{fig:RH_profiles} and figure \ref{fig:RH_profiles_biases}).