Solar flares are large-scale releases of energy in the solar atmosphere, which are characterised by rapid changes in the hydrodynamic properties of plasma from the photosphere to the corona. Solar physicists have typically attempted to understand these complex events using a combination of theoretical models and observational data. From a statistical perspective, there are many challenges associated with making accurate and statistically significant comparisons between theory and observations, due primarily to the large number of free parameters associated with physical models. This class of ill-posed statistical problem is ideally suited to Bayesian methods. In this paper, the solar flare studied by Raftery et al. (2009) is reanalysed using a Bayesian framework. This enables us to study the evolution of the flare's temperature, emission measure and energy loss in a statistically self-consistent manner. The Bayesian-based model selection techniques imply that no decision can be made regarding which of the conductive or non-thermal beam heating play the most important role in heating the flare plasma during the impulsive phase of this event.

Recurrent novae are binary stars in which a white dwarf accretes matter from a less evolved companion, either a red giant or a main-sequence star. They have dramatic optical brightenings of around 5-6 mag in V in less than a day, several times a century. These occur at variable and unpredictable intervals, and are followed by an optical decline over several weeks, and activity from the X-ray to the radio. The unpredictability of recurrent novae and related stellar types can hamper systematic study of their outbursts. Here we analyse the long-term lightcurve of RS Ophiuchi, a recurrent nova with six confirmed outbursts, most recently in 2006 February. We confirm the previously suspected 1945 outburst, largely obscured in a seasonal gap. We also find a signal via wavelet analysis that can be used to predict an incipient outburst up to a few hundred days before hand. This has never before been possible. In addition this may suggest that the preferred thermonuclear runaway mechanism for the outbursts will have to be modified, as no pre-outburst signal is anticipated in that case. If our result indeed points to gaps in our understanding of how outbursts are driven, we will need to study such objects carefully to determine if the white dwarf is growing in mass, an essential factor if these systems are to become Type Ia Supernovae. Determining the likelihood of recurrent novae being an important source population will have implications for stellar and galaxy evolution.

One way of revealing the nature of the coronal heating mechanism is by comparing simple theoretical one dimensional hydrostatic loop models with observations at the temperature and/or density structure along these features. The most well-known method for dealing with comparisons like that is the $\chi^2$ approach. In this paper we consider the restrictions imposed by this approach and present an alternative way for making model comparisons using Bayesian statistics. In order to quantify our beliefs we use Bayes factors and information criteria such as AIC and BIC. Three simulated datasets are analyzed in order to validate the procedure and assess the effects of varying error bar size. Another two datasets (Ugarte-Urra et al., 2005; Priest et al., 2000) are re-analyzed using the method described above. In one of these two datasets (Ugarte-Urra et al., 2005), due to the error estimates in the observed temperature values, it is not posible to distinguish between the different heating mechanisms. For this we suggest that both Classical and Bayesian statistics should be applied in order to make safe assumptions about the nature of the coronal heating mechanisms.