Riming detection using cloud radar Doppler spectra

Abstract:

Detecting riming using ground-based remote sensing (cloud radar) data has the potential to improve the prediction of aircraft traffic hazards and severe weather warnings. Deriving the rime mass fraction from measurements is also needed for observation-model comparison studies, which are an important step to evaluate and improve parameterizations of numerical weather prediction models. However, the retrieval of rime mass from ground-based remote sensing observations is complicated. Methods relying on the cloud radar mean Doppler velocity (MDV) are most common: Provided sufficient temporal and spatial averaging, a robust estimate of rime mass fraction can be obtained using the MDV. One limiting factor is however that periods with strong vertical air motions greatly influencing observed MDV have to be excluded from the data set. This impedes the application of this method to certain measurement locations in complex terrain, where orographically induced waves shift the observed MDV to more positive or negative values. Riming leads to a broadening of the cloud radar Doppler spectrum and an increase of its skewness, due to the presence of multiple hydrometeor populations within the same observation volume. We exploit this relation by defining a measure for the width of the Doppler spectrum, which is independent of orographically induced vertical air motion.

Using remote sensing and in-situ data from the ARM deployment in Hyytiälä, Finland, during the winter 2014/2015 as a starting point, we develop a novel riming criterion based on the cloud radar Doppler spectrum. Combining the Doppler spectrum width and the skewness, a threshold for classifying rimed spectra can be defined. We evaluate this criterion by comparison to the rime mass retrieved from particle imaging package (PIP) data, and by forward-modeling the PIP-measured size distributions into cloud radar Doppler spectra. We are using the Passive and Active Microwave Transfer forward operator (PAMTRA), utilizing a mix of T-Matrix and Self-Similar Rayleigh Gans scattering simulations to represent both rimed and unrimed particles. The novel riming detection technique is then applied to different cloud radar data sets obtained during field campaigns in Europe and Chile.