| الوصف: |
The mathematical modelling and numerical simulation of clouds and climate include numerous phenomena that are tough nuts to crack as they cover a wide range of spatiotemporal scales. In many ways, time is a vital factor, for instance, predicting the significance of a millisecond phenomenon for the future century is a major undertaking. Additionally, the computational time required by numerical models is a challenge. Luckily, we have a fine set of tools in our mathematical backpack. Here, we explore how a detailed cloud model can be improved to simulate the interactions with ice crystals. A new ice microphysics module is validated against a set of similar cloud models. Further on, the cloud model is shown to be an improvement over the previous generation of cloud models as it incorporates detailed aerosol-cloud interactions, which in our study is shown to impact cloud lifetime through ice nuclei recycling and marine ice nuclei import via updrafts. Additionally, the cloud model, which has a fine resolution in the order of meters, is harnessed to develop three different emulators to represent selected cloud processes in an improved detailed way. Emulators can be called also parametrisation or a machine learning model. Further on, created parameterisations are implemented within a global climate model, which has a much coarser resolution in the order of 10–100 kilometres. The implementation enables more precise climate simulations by having a more detailed subgrid scale description of cloud processes. As an adventurous side quest, we elaborate on how the proof-of-concept emulator could be embellished by showing an optimised way of creating the design of the simulation experiment in our applied case and we compare our results with the proof-of-concept method used in the study where the emulators were created. |