An Evaluation of SIP parametrizations in WRF in an Orographic Environment with Preliminary Data From CHOPIN

Mixed-phase clouds play a key role in weather and climate and are a major challenge to model accurately in models. The realistic representation of Mixed-phase clouds has been shown to be a large contributing factor to the uncertainty in the Coupled Model Intercomparison Project Phase 6 (CMIP6) model predictions of climate sensitivity and therefore understanding the capabilities and limitations of modelling mixed-phase cloud is of importance. Nevertheless, considerable progress on modeling microphysical processes in mixed-phase clouds, such as primary and Secondary Ice Production (SIP) has led to systematic improvements in model performance. In this thesis theWeather Research and Forecasting (WRF) model with the EPFL/FORTH cloud microphysical scheme improvements—referred to as WRF-SIP—(Georgakaki et al., 2024) was used to simulate weather, cloud formation, and precipitation events sampled during the Cleancloud Helmos OrograPhic sIte experimeNt (CHOPIN) campaign from the Fall of 2024 to Spring of 2025. CHOPIN takes place at Mount Helmos in Peloponnese, Greece, an ideal location for studying aerosol-cloud interactions in orographic mixed-phase clouds. The performance of SIP parameterizations in WRF are evaluated by (i) comparing model outputs to meteorological data from both a fixed weather station and radiosondes, (ii) comparing the model’s ability to capture boundary layer dynamics using atmospheric trace gasses as a proxy, and (iii) the model’s ability to capture cloud formation using reflectivity from a Ka-band (35 GHz) radar, by comparing the output of a forward operator (radar simulator) from the modeled cloud fields. Through the modelling of various days between October 2024 and December 2024, as well as a clustering analysis of the radar reflectivities, the capabilities of using SIP parameterizations in WRF are assessed with respect to particular cloud regimes. Through the use of in-situ meteorological data it was found that the MYNN2.5 closure boundary layer parameterization led to better prediction of simulated meteorological fields. The clustering analysis revealed 6 distinct cloud regimes and it was seen that the WRF-SIP model had better predictive quality on cloud regimes containing well developed and tall cloud structure. Lastly, it was seen that the cloud capturing capabilities of WRF without SIP and WRF-SIP were similar in most cases, but that WRF without SIP was better able to capture periods with fully glaciated high altitude clouds.