A highly efficient Voigt program for line profile computation

TL;DR

This paper introduces a highly efficient algorithm and Fortran90 implementation for calculating the Voigt function with high accuracy, significantly outperforming existing methods in speed, and suitable for large-scale and real-time scientific applications.

Contribution

The paper presents a novel, faster algorithm for Voigt function evaluation using Chebyshev polynomial approximation, improving computational efficiency over existing methods.

Findings

Algorithm achieves accuracy of 1.0e-6.

Significantly faster than existing algorithms.

Adopted in the Meudon PDR code at Paris Observatory.

Abstract

Evaluation of the Voigt function, a convolution of a Lorentzian and a Gaussian profile, is essential in various fields such as spectroscopy, atmospheric science, and astrophysics. Efficient computation of the function is crucial, especially in applications where the function may be called for an enormous number of times. In this paper, we present a highly efficient novel algorithm and its Fortran90 implementation for the practical evaluation of the Voigt function with accuracy in the order of 1.0e-6. The algorithm uses improved fits based on Chebyshev subinterval polynomial approximation for functions in two variables. The algorithm significantly outperforms widely-used competitive algorithms in the literature, in terms of computational speed, making it highly suitable for real-time applications and large-scale data processing tasks. The substantial improvement in efficiency positions the present algorithm and computer code as a valuable tool in relevant scientific domains. The algorithm has been adopted and implemented in the Meudon PDR code at Paris Observatory and is recommended for similar applications and simulation packages.