We place uncertainties on the derived model parameters
Here we show the best-fit model to the O VIII line
and two models that are marginally (at the 95% limit) consistent
with the data; they are the models with the highest
and lowest t* values possible.