Flare parameters inferred with 3D loop Models database

Valente Cuambe, Universidade Eduardo Mondlane

A solar flare is one of the most intense energetic phenomena in active regions in the solar atmosphere. Imaging of these regions in extreme ultraviolet and soft X-rays have revealed their complex magnetic configurations. During flares, the field geometry can be analyzed through its microwave component of emission with moderate image resolution. The main problems in this type of analysis are contained in the integral equations that describe both the spectra and images do not have analytical solutions. Therefore, only forward-fitting methods are possible to describe them. To work around this scenario, a database of flare models with about two hundred and fifty thousand elements was constructed to become a catalogue for flare analysis. The database was constructed using the geometry and physics of dipolar magnetic fields. The dipole model consists of a three-dimensional digitized region and spatial distribution of non-thermal electrons. The aim of the database is to speed up the search for the parameters in a flare, using pre-calculated models with forward-fitting methods. On the other hand, a flare catalogue was built (reduced database ) based on Nobeyama polarimeter (NoRP) flux densities and NoRH brightness maps observations (but not restricted to) including known general properties of a solar flare. Given images, the flare position and flux densities in four used frequencies in the database it returns the best model representation. We analysed two different methods to search for models in the database using a chi-squared and the weighted mean of one hundred best models in the database. As a result, we found that about ~80 percent of the ten analyzed parameters of 1000 simulated flares were recovered with the relative error <20 percent in average. From the statistic analyses of the NoRH flares, now using this database of non-homogeneous models, we found the following results: the distribution of the energy spectral index delta peaks ~3, nonthermal electrons density tending to be lower than <10^7 cm^-3, and the peak of the magnetic field distribution B_photosphere > 2000 G. We note some preferences for extended loops with height as greater than ~ 2.6x10^9 cm and looptop events. We conclude that the use of this database with a moderate number of elements increases the possibility to find good results during a flare analysis and explore some statistical properties of flares. On the other hand, it also speeds up the search for the parameters in a flare with a high level of acceptance.