Accurate fitting formulae to the synchrotron function, F（x）, and its com- plementary function, G（x）, are performed and presented. The corresponding relative errors are less than 0.26% and 0.035% for F（x） and G（...Accurate fitting formulae to the synchrotron function, F（x）, and its com- plementary function, G（x）, are performed and presented. The corresponding relative errors are less than 0.26% and 0.035% for F（x） and G（x）, respectively. To this end we have, first, fitted the modified Bessel functions, Ks/3（x） and K2/3（x）. For all the fitted functions, the general fit expression is the same, and is based on the well known asymptotic forms for low and large values of z for each function. It consists of multi- plying each asymptotic form by a function that tends to unity or zero for low and large values of z. Simple formulae are suggested in this paper, depending on adjustable parameters. The latter have been determined by adopting the Levenberg-Marquardt algorithm. The proposed formulae should be of great utility and simplicity for com- puting spectral powers and the degree of polarization for synchrotron radiation, both for laboratory and astrophysical applications.展开更多
基金supported by the National Administration of Scientific Research NASR-DZ, of Algeria, in the framework of National Projects of Research (NPR)
文摘Accurate fitting formulae to the synchrotron function, F（x）, and its com- plementary function, G（x）, are performed and presented. The corresponding relative errors are less than 0.26% and 0.035% for F（x） and G（x）, respectively. To this end we have, first, fitted the modified Bessel functions, Ks/3（x） and K2/3（x）. For all the fitted functions, the general fit expression is the same, and is based on the well known asymptotic forms for low and large values of z for each function. It consists of multi- plying each asymptotic form by a function that tends to unity or zero for low and large values of z. Simple formulae are suggested in this paper, depending on adjustable parameters. The latter have been determined by adopting the Levenberg-Marquardt algorithm. The proposed formulae should be of great utility and simplicity for com- puting spectral powers and the degree of polarization for synchrotron radiation, both for laboratory and astrophysical applications.
