A decision tree algorithm for calculating dielectronic recombination rates in low temperature astrophysical plasmas*

Dielectronic recombination (DR) is one of the dominant mechanisms for electron recombination in low temperature astrophysical plasmas. In this work, we address the issue of low temperature DR rate coefficients for astrophysical photoionized plasmas, where DR rates are known to have large uncertainties in their calculated values. It can be seen from discrepancies with storage ring measurements that current theoretical methods have large uncertainties in the low temperature regime due to uncertainties in calculating low-n doubly excited states. AUTOSTRUCTURE takes the multi-configuration Breit-Pauli approach and is a commonly used code for computing DR rates. The uncertainties can be diminished with larger configuration sets but, as the size of the selection grows, the problem of selecting an optimum set quickly becomes non-trivial. The Complex Resolved Ion Spectroscopy Tree Algorithm (CRISTAL) is a decision tree algorithm we have designed to efficiently search the space of possible configuration sequences and compare calculated energies with existing storage ring data to optimize the selection. Using CRISTAL in combination with AUTOSTRUCTURE, we have reduced the uncertainties in low energy resonance positions to within roughly 0.1 eV in comparison with existing storage ring measurements for Ne3+ recombination. We test these new ion models against new measurements that are being performed at the heavy ion storage ring CRYRING@ESR at the FAIR facility in Darmstadt, Germany.