Linda Schlemmer's research while affiliated with Deutscher Wetterdienst and other places

... Traditionally, dynamical cores and physical parameterizations have been engineered in isolation for the sake of tractability (Donahue and Caldwell (2018); Gross et al. (2018), and references therein). These two independent components are coupled and advanced using the same time step, either parallel or sequentially split (Ubbiali et al., 2021). Ubbiali et al. (2021) analyzed six strategies of dynamics-physics coupling in atmospheric models. ...

... The grid has 65 vertical levels and the horizontal grid spacing of about 2 km. ICON in the large eddy mode (LEM; Bašták Ďurán et al., 2021;Dipankar et al., 2015;Heinze et al., 2017) was used as a reference. The ICON-LEM was run in a limited-area setup similar to ICON-LAM, but forced by the ICON-D2 operational forecasts and with an LEM physics setup (Dipankar et al., 2015). ...

... Another factor in computational cost is the need to run atmospheric models for climate projections using large ensembles at relatively coarse horizontal resolutions, in which mountains are poorly resolved. Many recent studies have demonstrated that the improvements seen in the troposphere when horizontal resolution is increased is dominated by the increase in the resolved orographic detail (Kanehama et al. 2019; Berckmans Heim et al. 2020). What this implies is that there is significant potential for increased skill from improvements in parametrization of orographic processes at coarser resolutions. ...

... With grid spacings of a few kilometers, these models start to resolve deep convection explicitly. Most GSRMs therefore run without convective parameterization , though there is ongoing discussion about whether a scale-aware convective parameterization might improve certain model skills (e.g., Becker et al., 2021;Vergara-Temprado et al., 2020). One big advantage of resolving deep convection explicitly is that small-scale turbulent mixing and microphysical processes are directly coupled to the resolved circulation, rather than being partly represented by the convective parameterization. ...

... A possible cause for the weakness of the surface flux feedback at later stages in the medium-f simulations lies in the model design. Mol et al. (2019) note that many CRMs employ a minimum wind speed threshold to compute latent and sensible heat fluxes. In SAM, this wind speed is 1 m s −1 , such that any grid point with a weaker wind is artificially raised to that threshold when calculating fluxes. ...

... hourly) climate models for downscaling have coarse temporal resolution (Dyrrdal et al., 2018;Iles et al., 2020) making it difficult to precisely analyse urban flood risks associated with climate change. Although recent climate models allowing for explicit modelling of deep convection (Prein et al., 2015) can 45 simulate 5 minute precipitation fields (Meredith et al., 2020), these products will not be available on a large scale in the near future due to computational and data storage limitations (Schär et al., 2020). ...

... First, wind patterns are more uncertain in new locations that have shorter historical observation records. Second, new sites may have lower quality wind, with characteristics that are more difficult to incorporate into existing forecasting methods, such as terrain complexity [31][32][33][34] and variable land use (e.g., urban environments). ...

... Climate modelers place great confidence in the next generation of kilometer-scale (or storm resolving) climate models (Neumann et al., 2019;Stevens et al., 2019), which will allow explicit representation of deep convection and more detailed representation of the bulk influence of diabatic processes on the adiabatic flow, even if structural details of individual clouds remain unresolved (Palmer, 2014;Prein et al., 2015;Schär et al., 2020). Regional kilometer-scale simulations have indeed improved the representation of, for example, the diurnal cycle of convection and precipitation (Panosetti et al., 2018(Panosetti et al., , 2019Stevens et al., 2020) as well as the structure of mesoscale eddies along cold fronts (B. Harvey et al., 2017;, tornadoes (Hanley et al., 2016), cloud cover (Hentgen et al., 2019), and the response of low-lying clouds over the tropics to warming (Schneider et al., 2019). ...

... Also, the impact of the different sizes of mountains on error growth or predictability has been unexamined. The mountain's height or width impacts the development of moist convection and the accompanying rainfall amount and distribution (Flesch and Reuter 2012;Imamovic et al. 2019;Mulholland et al. 2020). A study should assess whether this sensitivity to mountain geometry is reflected in the error growth or predictability limit. ...

... However, models should be further developed in order to constrain the range of projections, and to make their output more directly applicable to agricultural impact studies. The surface and soil hydrology of the models is of particular concern in this context (Schlemmer et al 2018). Although the model ensemble can qualitatively reproduce observed soil moisture variability and its annual cycle (Knist et al 2017), models differ substantially in the parameterizations, the maximum depth and field capacity. ...