deletions | additions       diff --git a/scattering part2.tex b/scattering part2.tex  index 75d6168..255d0b0 100644  --- a/scattering part2.tex  +++ b/scattering part2.tex 
...
The scattering timescale can also be expressed in terms of dispersion measure (DM). Here we will use the empirical model of \citet{Bhat_2004}, that relates $\tau_s$ to DM via  \begin{equation}  \label{eq:scat}  \log \tau_s = -6.46 + 0.154 \log_{10} ({\rm DM})+ \log{\rm DM}+  1.07 (\log{\rm DM})^2 - 3.86 \log{\nu}. \end{equation}  where $\tau_s$ is the scattering time in ms and $\nu$ is the observing frequency in GHz. The scatter of measured values about this relation ship has a standard deviation of several orders of magnitude. In the remainder of this document we take a conservative approach and assume scattering times 4 orders of magnitude smaller than those suggested by the above relation, i.e. the best-case scattering. As we will see, this approach results in reasonable requirements on the SKA1 pulsar timing instrumentation.   
...