Evaluation of Voigt algorithms for the ISS/JEM/SMILES L2 data processing system

We present a calculation of the accuracy and speed of the Voigt function of various algorithms (Armstrong (1967), Hui et al. (1978), Humlicek (1982) and improved version of Kuntz (1997) developed in this study) for making the faster and higher accuracy forward model required for the Superconducting Submillimeter-Wave Limb-Emission Sounder (SMILES) Level 2 (L2) data processing system.The maximum relative error of the improved Kuntz’s and Armstrong’s algorithm is ϵmax∼2.52×10-5ϵmax∼2.52×10-5 in the high-altitude range (y<1)(y<1) and a minimum value of ϵmax∼9.38×10-5ϵmax∼9.38×10-5 compared to the other results in the mid- and low-altitude range (y>1)(y>1), which meet a required precision of the SMILES L2 data processing system. The improved algorithm also shows the fastest calculation time compared to the other algorithms over the observable altitudes of SMILES. It is therefore concluded that the improved algorithm should be applied for the SMILES L2 data processing system.