Motivation for the visit

The McSnow Lagrangian cloud model is one of the pivotal tools and methods used for scientific research in project PRISTINE (University of Cologne). Intensive McSnow simulations are performed to obtain detailed snow particle properties, which are used to further simulate snow shapes. On this ground, PRISTINE leverages continuous fruitful collaboration with the leading developers of McSnow at Deutscher Wetterdienst (DWD) with Dr. Christoph Siewert and Dr. Axel Seifert.

McSnow Model

The McSnow model was developed by Brdar and Seifert (Brdar and Seifert, 2018). McSnow is a (1D box model) Lagrangian particle model that simulates the evolution of ice particles via depositional growth, aggregation, riming, melting, and warm phase processes. To reduce computational effort, a super droplet method (Shima et al., 2009) has been used in McSnow for the evolution of multi-property ice particles. Each super-particle represents multiple real particles with the same attributes and position, and each super-particle has a multiplicity (number of real particles represented by it). Note here that no two real particles (true ice particles) have exactly the same position and attributes, and in this sense, a super-particle is a kind of coarse-grained view of particles both in real space and attribute space. McSnow is a 1D box model with no horizontal resolution. We consider a five-square-meter grid-box area with several height layers inside the McSnow simulation. Thus, a super-particle is more of a representation of an ensemble of real particles within a given volume. The super-particle’s evolution is then tracked to get aggregation history. In McSnow, ice mass, rime mass, volume, number of monomers, and shape of the monomers are tracked for each super-particle. The model has been improved during the PROM-I project IMPRINT to explicitly model frozen particle shapes and shape-dependent ice processes. Modeling of ice particle structures with a much higher level of detail is required for advanced scattering simulations in project PRISTINE.

McSnow Simulation for Real Case Scenario

Our objective of realistic snow shape simulation using the McSnow and snow particle model combination includes real test cases from the winter TRIPEX Campaign (2018-19) over Julich. We selected some of the campaign days where there were snowfall events. Radar reflectivity images from the campaign observations help us look into the cloud patterns, and we choose cloud cases that are stratiform in nature to avoid heterogeneity. As input to the McSnow model, that particular cloud event's thermodynamic profile (pressure, temperature, height, and specific humidity) must be provided. We use thermodynamic profiles from ICON simulations (data storage at the University of Cologne). There is a challenge when we provide ICON simulated profiles to the McSnow. ICON is a 3D model that uses a 2-mom microphysics scheme, while McSnow is a 1D Lagrangian model. Therefore, it is essential to check whether the ice and snow profiles simulated in McSnow are comparable to those of ICON ice and snow profiles. A deeper understanding was needed to choose the TRIPEX campaign dates or events for the McSnow simulation. During the one-week visit to DWD Offenbach, I discussed the methodologies and steps appropriate for ICON and McSnow comparisons with Dr. Christoph Siewert (DWD) and Dr. Axel Seifert (DWD). Also, I simulated McSnow ice and snow profiles to check which dates of the TRIPEX campaign are unsuitable for MCSnow simulations and which are appropriate. For example, figure 1 below shows ice number concentration (QNI), ice mixing ratio (QI), snow number concentration (QNS), and snow mixing ratio (QS) simulated from the McSnow Lagrangian model (blue line) for the date 21st December’2018 and time 07 am and the ICON simulated profiles (Orange line). Besides that, McSnow also includes a two-moment scheme (same as ICON) for the simulation of ice and snow profiles. Therefore, we have also analyzed the two-moment simulation from McSnow (Green line). Figure 1 shows relevant matches between ICON and MCSnow ice and snow profiles. Therefore, we choose this real case from the TRIPEX campaign. The initial conditions used for the McSnow simulation are written on the upper left of Figure 1.

Summary

During the five days of a research visit to DWD, I was introduced to the tool how to select real cases for McSnow simulation, which is an essential aspect of our project. I will use those tools to choose more real cases for the McSnow simulation.

3. Case study results

A case study comparing Doppler spectra and Ze from the 22nd January 2022 collected during TRIPEx-pol campaign (von Terzi et al. 2022) reveals a generally good agreement between observations and the reconstruction from the retrieved (micro-) physical properties. A closer look identifies areas of small discrepancies: the Doppler spectra shown Fig. 1 show a slight underestimation of the main peak (slightly too small sZe at ranges larger than approx. 3000m). While the increase of spectral width at approx. 3000m is captured by the retrieval, the second mode observed at heights between 2000m and 3000m is not represented correctly. The underlying representation of the PSD as a gamma distribution is not able to produce a second mode. A similarly good agreement can be seen when comparing several hours (shown in Fig. 2) of Ze observations and reconstructions.