The k-bin tool: Fast and flexible k-distribution algorithms written in Python

Radiative transfer simulations (RTS) still face significant challenges in accurately representing the highly complex gas absorption spectra of the Earth’s atmosphere. Line-by-line RTS achieves high accuracy by solving radiative transfer equations for narrow spectral intervals, but at a considerable computational cost. Especially in remote sensing and climate modeling, a trade-off between efficiency and accuracy must be done. k-distribution methods are widespread in the scientific community and offer a way to make this trade-off. k-distribution methods reorder the absorption spectra k for a given spectral interval and find appropriate so-called k-bins. In the k-space much less integration points can be used, while maintaining high accuracy. The way to find optimal k-bins differs from method to method and depends on the application. In this paper, we present the flexible and fast k-bin tool. The python based lightweight k-bin tool provides a variety of different k-distribution methods and configuration options. One k-distribution method is the in-house developed k-bin approach. The different setups of the tool can be easily compared, helping to decide which method and configuration is best suited for a given application. We encourage the user of the tool to continue to optimize the k-bin tool and to extend it with new approaches and functionalities. • Lightweight, fast, flexible, and extendable Python-based k-bin tool • Support of various k-distribution methods • Addressing challenges in accurately representing gas absorption spectra in atmosphere • Support in decision making for the trade-off between efficiency and accuracy of RTS • Easy comparison of different setups to determine the optimal method and configuration • Illustration of advantages and disadvantages of the in-house developed k-bin approach