Modeling & Machine Learning
Modeling
As a part of the HighResMountains, two high-resolution data sets are compared - dynamically downscaled km-scale simulations versus ÖKS statistically downscaled i.e. bias adjusted data. The main goal is to assess how are extreme events, processes related to them and their changes with further warming, presented in these datasets. [PI: N. Ban]
As a part of the kmMountains project, COSMO simulations are performed with a horizontal grid spacing of 2.2 and 1.1 km over two mountainous regions for multi-decadal periods in the present and future climate. The main goal is to evaluate the performance of the high-resolution models over complex topography and assess how furter warming of the atmosphere will affect the mountain climate. [PI N. Ban]
As part of the ASTER project WRF simulations with a grid spacing of 1 km are performed for the Inn Valley to evaluate the impact of the land-use dataset, boundary-layer parameterizations, initial soil fields, and the model evaluation strategy on the model’s representation of the surface-energy balance and the valley atmosphere. [UIBK PI: M. Lehner]
A subgroup of the MoBL WG co-led by A. Gohm is performing an intercomparison of mesoscale models for a case study of thermally driven flows in the Inn Valley.
A subgroup of the MoBL WG led by M. Lehner is performing an intercomparison of mesoscale models for a case study of cold-air pool formation in the Inn Valley.
Several case studies at kilometer and hectometer resolution are performed within four master’s theses supervised by A. Gohm to study the impact of model resolution, boundary layer parameterization and tributary valleys on the representation of MoBL processes.
As part of the SCHiRM project, building-resolving simulations will be conducted for the Innsbruck Atmospheric Observatory using PALM. Model output will be compared with observations and the model will be used to study the urban and orographic effects on atmospheric conditions in the city. Mesoscale model performance (including WRF with the multi-layer urban canopy parameterisation) will also be considered. [PI H. Ward]
As part of the Unicorn project, the group led by I. Stiperski will perform ultra-high resolution LES and DNS to examine the influence of complex terrain on turbulence characteristics and surface-exchange processes. [PI: I. Stiperski]
Machine Learning
As part of the Unicorn project, the group led by I. Stiperski will perform ultra-high resolution LES and DNS to examine the influence of complex terrain on turbulence characteristics and surface-exchange processes. [PI: I. Stiperski]
One of the major challenges in machine learning is data engineering and data-driven model discovery. Developing methods that objectively extracts useful features from complex data is the primary focus of the ISM project. [PI: K. Lapo]