Final text corrections

dual-problems.tex

@@ -630,7 +630,7 @@ \subsection{Gibbous ice shelf} \label{sec:gibbous-ice-shelf}
 
 We repeated this phase of the experiment using a comparable solver for the primal problem.
 When we use no minimum thickness at all, the solver for the primal form diverges as soon as there are any ice-free areas.
-To remedy this problem, we clamped the thickness from below at 1mm.
+To remedy this problem, we clamped the thickness from below at 1 mm.
 Figure \ref{fig:gibbous-residuals} shows the number of Newton iterations necessary to obtain the desired level of convergence through two calving events using both the primal and dual forms.
 In each case, the number of iterations goes up after a calving event.
 As the system relaxes back, the number of iterations decreases again.
@@ -684,7 +684,7 @@ \subsection{Kangerlussuaq Glacier}
 We ran several instances of the experiment outlined above with different values of the maximum melt rate $m_0$.
 In general, the total volume of ice in the simulated domain oscillates from summer lows to winter highs over a wide range of $m_0$ values.
 With too low or too high a maximum melt rate, there is an additional secular trend in the volume time series as the glacier advances or retreats down the fjord.
-We found that taking $m_0$ on the order of 30 km/yr makes the yearly-averaged volume roughly constant; see figure \ref{fig:kangerd-volumes}.
+We found that taking $m_0$ on the order of 30 km yr${}^{-1}$ makes the yearly-averaged volume roughly constant; see figure \ref{fig:kangerd-volumes}.
 Spread over an inland distance of roughly 1 km in a 5 km-wide fjord for only the summer season, this gives a total discharge roughly of the same order as the observed value of 24 km${}^3$ yr${}^{-1}$ \citep{king2018seasonal}.
 
 Figure \ref{fig:kangerd-contours} shows the evolution of the calving terminus from a minimum to the following maximum extent.