Order statistics and energy-ordered histograms: an exploration into continuum gamma spectroscopy

Full text: Let us consider the gamma rays emitted in a fusion-evaporation heavy ion collision. We have different ways to organize and analyze the detected gamma rays: coincidences, angular correlations, DCO, lineshapes, etc. With these methods a wide variety of phenomena mainly in the discrete regime has been investigated. On the opposite, the region of the continuum, though rich in new physics has been by far not as well examined experimentally. The reason for this situation is -in contrast to the discrete spectroscopy- a lack of experimental methods. An approach towards continuum gamma spectroscopy is the so called (H)-(k) technique [1]. Taking advantage of the fact that the gamma rays of an event-by-event detected gamma cascade can be off-line ordered according to their energies, the method of Energy-Ordered Spectra (EOS) was proposed [2] with the goal of picking up the gamma radiation emitted in defined regions of high intrinsic energy and spin. This procedure's success would allow the extension of discrete spectroscopy methods into continuum spectroscopy. Monte Carlo generated EOS show that a judicious comparison between them and experimental ones would allow to study the eventual variation of the level density parameter as a function of intrinsic excitation energy and/or spin [3] ,141. The same work showed that there is even an analytical way to obtain EOS from defined regions of the phase space E - I. The analytical approach comes from the application of the mathematical theory of Order Statistics [5]. This represents a clear improvement to comparisons with numerical simulations: the analytical (fitting) function involves a reduced number of parameters reflect- ing the specific physics under study whereas the large number of input values necessary in a gamma decay simulation does not allow straightforward interpretations. In the present work it is shown that the artificial discontinuities appearing in the previously [3] found fitting function can be avoided. The potential use of this function in the analysis of experimental EOS is then discussed. (Author)