Modeling the Phase Transition Associated with Melting Snow in a 1D Kinematic Framework: Sensitivity to the Microphysics

Abstract

Abstract-A simple 1D kinematic cloud model coupled to a two-moment bulk microphysics scheme is used to perform quasiidealized simulations of snow, with a prescribed upper boundary snow field based on observed radar reflectivity and temperature, falling into a low-level melting layer. The model realistically simulates the formation of a nearly isothermal layer below the melting level, the surface precipitation rate, and the phase transition from liquid to solid, consistent with observations for this case. A series of test runs is performed to examine the sensitivity of modeling the timing and duration of the phase transition period to details of specific parameterization aspects related to snow in the microphysics scheme. The sensitivity tests include varying the number of prognostic moments, the mass-diameter relation, the fall velocity-diameter relation, the treatment of aggregation, and the lower limit for the slope of the size distribution. It is shown that the simulated transition period, for such a case with the initial melting level being close to the surface, can be quite sensitive to model parameters specified within realistic ranges and/or ranges within our physical understanding.

Impact of melting snow on the valley flow field and precipitation phase transition

“…While this has a notable but temporary effect on the valley flow structure, it appears that convective overturning has little overall effect on the rate at which the valley cools. This is consistent with the fact that the two-dimensional simulation results are similar, in terms of temperature changes and precipitation phase transition, to the one-dimensional results of Milbrandt et al (2014), where convection is absent.…”

“…The control run and the sensitivity tests with different prescribed precipitation rates (described below) were done using a two-moment bulk microphysics scheme as described in Milbrandt and Yau (2005) (hereafter referred to as MY2). Given the sensitivity to the parameterization of snow shown in Milbrandt et al (2014), it should be noted that the original version of the scheme has been used in this study.…”

“…Therefore, it was assumed that snow is falling continuously from above the melting layer to yield a surface precipitation rate of approximately 5 mm h −1 . The snow field was initialized with a mixing ratio, q s = 1.25 g kg −1 , and total number concentration, N s = 9860 m −3 , at the model level corresponding to an elevation of 2.3 km, based on observed radar reflectivity and temperature, assuming the relation between the snow intercept size distribution parameter and temperature from Thompson et al (2004) (see Milbrandt et al, 2014) for details).…”

“…The sensitivity to microphysical assumptions related to snow on the diabatic cooling effects and the resulting precipitation phase changes were examined in Milbrandt et al (2014) in a simple one-dimensional framework for the 13-14 February 2010 case in the Whistler area. The snow quantity aloft corresponding to radar observations was prescribed with an observed temperature and humidity profile with melting and cooling rates simulated with a bulk microphysics scheme.…”

Three‐Moment Representation of Rain in a Bulk Microphysics Model

“…The 1‐D kinematic driver model from Milbrandt et al () is used here. The model consists of 41 vertical levels which extend up to 12.8 km with a vertical grid spacing between 35 and 345 m at the relevant altitudes.…”