The predictions of general-circulation models (GCMs) are sensitive to the assumed cloud overlap within a vertical column of model grid boxes, but until now no reliable observations of the degree of cloud overlap have been available. In this note we derive the overlap characteristics of clouds from 71 days of high vertical resolution 94 GHz cloud radar data in the UK. It is found that, contrary to the assumption made in most models, vertically continuous clouds tend not to be maximally overlapped. Rather, the overlap of clouds at two levels tends to fall rapidly as their vertical separation is increased, and for levels more than 4 !un apart, overlap is essentially random. A simple inverse-exponential expression for the degree of overlap as a function of level separation is proposed that could, once results become available from a variety of other locations and seasons, be implemented in current GCMs with relatively little difficulty.

Retrievals of ice and snow are made from Ka‐ and W‐band zenith‐pointing Doppler radars at Hyytiälä, Finland, during the snow experiment component of the Biogenic Aerosols: Effects on Clouds and Climate (2014) field campaign. In a novel optimal estimation retrieval, mean Doppler velocity is exploited to retrieve a density factor parameter, which modulates the mass, shape, terminal velocity, and backscatter cross sections of ice particles. In a case study including aggregate snow and graupel we find that snow rate and ensemble mean ice density can be retrieved to within 50% of in situ measurements at the surface using dual‐frequency Doppler radar retrievals. While Doppler measurements are essential to the retrieval of particle density, the dual‐frequency ratio provides a strong constraint on particle size. The retrieved density factor is strongly correlated with liquid water path, indicating that riming is the primary process by which the density factor is modulated. Using liquid water path as a proxy for riming, profiles classified as unrimed snow, rimed snow, and graupel exhibit distinct features characteristic of aggregation and riming processes, suggesting the potential to make estimates of process rates from these retrievals. We discuss the potential application of the technique to future satellite missions.

No abstract

Abstract. The accurate representation of ice particles is essential for both remotely sensed estimates of clouds and precipitation and numerical models of the atmosphere. As it is typical in radar retrievals to assume that all snow is composed of aggregate snowflakes, both denser rimed snow and the mixed-phase cloud in which riming occurs may be under-diagnosed in retrievals and therefore difficult to evaluate in weather and climate models. Recent experimental and numerical studies have yielded methods for using triple-frequency radar measurements to interrogate the internal structure of aggregate snowflakes and to distinguish more dense and homogeneous rimed particles from aggregates. In this study we investigate which parameters of the morphology and size distribution of ice particles most affect the triple-frequency radar signature and must therefore be accounted for in order to carry out triple-frequency radar retrievals of snow. A range of ice particle morphologies are represented, using a fractal representation for the internal structure of aggregate snowflakes and homogeneous spheroids to represent graupel-like particles; the mass–size and area–size relations are modulated by a density factor. We find that the particle size distribution (PSD) shape parameter and the parameters controlling the internal structure of aggregate snowflakes both have significant influences on triple-frequency radar signature and are at least as important as that of the density factor. We explore how these parameters may be allowed to vary in order to prevent triple-frequency radar retrievals of snow from being over-constrained, using two case studies from the Biogenic Aerosols – Effects of Clouds and Climate (BAECC) 2014 field campaign at Hyytiälä, Finland. In a case including heavily rimed snow followed by large aggregate snowflakes, we show that triple-frequency radar measurements provide a strong constraint on the PSD shape parameter, which can be estimated from an ensemble of retrievals; however, resolving variations in the PSD shape parameter has a limited impact on estimates of snowfall rate from radar. Particle density is more effectively constrained by the Doppler velocity than triple-frequency radar measurements, due to the strong dependence of particle fall speed on density. Due to the characteristic signatures of aggregate snowflakes, a third radar frequency is essential for effectively constraining the size of large aggregates. In a case featuring rime splintering, differences in the internal structures of aggregate snowflakes are revealed in the triple-frequency radar measurements. We compare retrievals assuming different aggregate snowflake models against in situ measurements at the surface and show significant uncertainties in radar retrievals of snow rate due to changes in the internal structure of aggregates. The importance of the PSD shape parameter and snowflake internal structure to triple-frequency radar retrievals of snow highlights that the processes by which ice particles interact may need to be better understood and parameterized before triple-frequency radar measurements can be used to constrain retrievals of ice particle morphology.