Taranis: A New Framework for Physically Constrained Radar Processing

Abstract:

The measurements obtained from weather radar are often not suitable for atmospheric cloud and precipitation process studies without additional processing that transforms these quantities into geophysically meaningful parameters. Taranis is a set of new dual-polarization weather radar processing algorithms designed to work with each other to convert raw radar measurements into useful geophysical quantities for observation and modeling studies. It is designed to work across multiple radar wavelengths in a variety of weather regimes.

The core of Taranis is a new physically constrained specific differential phase (K_DP) estimation algorithm that works at multiple frequencies combining the best of the two most common algorithms, linear programming and spline based adaptive filtering, allowing for high-resolution K_DP retrievals. The results from this processing are then passed into a new attenuation correction algorithm based on the self-consistency of dual-polarization measurements based on T-Matrix scattering. After correction, various microphysical quantities such as rain-rate, drop size distributions, and hydrometeor identification are estimated. Supporting all of this is a robust set of texture-based data quality masks. All algorithms are implemented using a mixture of Python for the interfaces, with C implementations under the hood for computational performance.

This work will discuss the Taranis framework and new algorithms, and demonstrate the benefit of the increased K_DP accuracy and spatial resolution offered by these new approaches. We will show applications of Taranis primarily to observations collected during the 2018-2019 Department of Energy Cloud, Aerosol, and Complex Terrain Interactions (CACTI) field campaign in Argentina, but also draw examples from other campaigns and radars operating at a variety of wavelengths in different weather regimes.