Description of the project

The WG “Atmospheric Dynamics” makes the largest contribution from KIT to “Waves to Weather”. Peter Knippertz is coordinator for the location Karlsruhe, Andreas Fink is coordinator of Reseach Area C – “Predictability of local weather”.

The WG is involved in five sub-projects:

• B6 - New data assimilation approaches to better predict tropical convection
  (PIs: Janjic-Pfander, Tijana; Hoose, Corinna; Knippertz, Peter)

  Weather prediction in the tropics is challenging. One reason are large errors in our best estimates of the atmospheric starting conditions for weather forecasting, which are created by merging a forecast and observations through so-called data assimilation. Here, we will investigate whether we can improve on the so far ill-represented coupling between large-scale tropical waves and rainfall through higher model resolution, a more realistic representation of clouds or better data assimilation algorithms that include energy conservation and other physical constraints.

• C2 - Statistical-dynamical forecasts of tropical rainfall
  (PIs: Fink, Andreas; Gneiting, Tilmann; Knippertz, Peter)

  Precipitation forecasting in the tropics is one of the remaining large challenges in weather prediction today. The interdisciplinary project aims at building novel statistical-dynamical forecast systems for northern tropical Africa and for the entire tropical belt. Particular attention will be paid to large-scale tropical wave phenomena, as we expect these to exert a relatively predictable control on rainfall variability on timescales of days to weeks. The new methods will be compared to established forecast methods to assess their added value.

• C3 - Predictability of tropical and hybrid cyclones over the North Atlantic Ocean
  (PIs: Fink, Andreas; Riemer, Michael; Schömer, Elmar)

  This interdisciplinary project investigates the predictability of tropical and tropical-like cyclones over the North Atlantic Ocean. Beyond seven days, a statistical-dynamical forecast approach is used employing neural networks to predict bulk properties of cyclones. Up to ten days, the causes of sudden changes in the ensemble forecast statistics with lead times for a larger number of individual storms will be studied. Geometry-based, low-dimensional descriptions of atmospheric wave disturbances will facilitate the necessary application of analysis methods to large forecast ensembles.

• C4 - Predictability of European Heat Waves
  (PIs: Fink, Andreas; Wirth, Volkmar)

  This project investigates the predictability of European heat waves, which gets increasing relevance in the light of the projected global warming. In phase 1 we illuminated the different physical processes playing a role for the establishment and persistence of recent heat waves. During phase 2 we plan to (1) put these results on a firmer basis by extending the methods used and the forecast lead times considered, and (2) explore the implications of our previous results for forecasting European heat waves using a variety of methods.

• C5 - Dynamical feature-based ensemble postprocessing of wind gusts within European winter storms
  (PIs: Knippertz, Peter; Lerch, Sebastian)

  Winter storms are among the most significant natural hazards in central Europe. Here, we develop objective identification algorithms for individual high-wind regions within winter storms using sophisticated visual analysis techniques. We will then analyze the dynamic behavior and evaluate forecast performance and error characteristics for each region. Using modern machine learning approaches, postprocessing methods based on the statistical behavior in the past will finally be developed to correct predictions for the future and to estimate forecast uncertainty, ensuring physical consistency.

The WG “Cloud Physics” is involved in four sub-projects of “Waves to Weather”:

• B1 - Microphysical uncertainties in hailstorms using statistical emulation and stochastic cloud physics
  (PIs: Hoose, Corinna; Kunz, Michael; Miltenberger, Annette)

  Severe convective storms, especially if associated with large hail, frequently cause substantial damage. When and where hailstorms occur and how intense they are, is very difficult to forecast. In this project, we use advanced statistical methods to investigate to what extent the forecast quality and reliability can be improved by reducing uncertainties in the numerical description of clouds, taking into account the additional uncertainty in the knowledge of relevant environmental parameters such as temperature, humidity and aerosol concentration.

• B3 - Sources of uncertainty for convective-scale predictability
  (PIs: Barthlott, Christian; Keil, Christian)

  Forecasting convective precipitation remains one of the key challenges in numerical weather prediction. The overall goal of this project is to determine how key physical processes contribute to uncertainty in forecasts of convective precipitation. Using the new convective-scale forecasting system ICON, we will adress individual and collective effects of soil moisture and microphysical uncertainties (e.g. drop size distribution) on precipitation predictability. Using a comprehensive ensemble data assimilation and forecasting system allows a seamless investigation of their impact on process uncertainties. Furthermore, the role of orography in controlling land surface–atmosphere and aerosol–cloud interactions is addressed.

• B6 - New data assimilation approaches to better predict tropical convection
  (PIs: Janjic-Pfander, Tijana; Hoose, Corinna; Knippertz, Peter)

  Weather prediction in the tropics is challenging. One reason are large errors in our best estimates of the atmospheric starting conditions for weather forecasting, which are created by merging a forecast and observations through so-called data assimilation. Here, we will investigate whether we can improve on the so far ill-represented coupling between large-scale tropical waves and rainfall through higher model resolution, a more realistic representation of clouds or better data assimilation algorithms that include energy conservation and other physical constraints.

• B8 - Role of uncertainty in ice microphysical processes in warm conveyor belts
  (PIs: Grams, Christian; Hoose, Corinna; Miltenberger, Annette)

  In the warm sector of an extratropical cyclone, a rapidly ascending air stream - the warm conveyor belt - transports air from low levels to the tropopause. It modifies the flow in the upper troposphere and the evolution of downstream weather. However, the numerical description of clouds in the warm conveyor belt, which contain liquid water and ice, is still uncertain. Here we will investigate how the uncertainty of ice microphysical processes propagates to the warm conveyor belt outflow. For this, we will use nested simulations with ICON, creating a large ensemble and evaluate it by means of Lagrangian diagnostics and statistical emulation.

The WG “Large-scale Dynamics and Predictability” is involved in two sub-projects of “Waves to Weather”:

• A8 - Dynamics and predictability of blocked regimes in the Atlantic-European region (Pis: Grams, Christian; Riemer, Michael; Wirth, Volkmar)

  Blocking high-pressure systems are important for atmospheric predictability at lead times beyond one week because they bring persistent weather conditions to specific regions, potentially leading to high-impact weather events. The correct representation of the onset and decay of these blocking systems, however, is a major challenge for numerical weather prediction models. This project thus investigates the processes that govern the evolution and associated forecast errors of blocking high-pressure systems to better understand under which conditions these systems lead to reduced and enhanced atmospheric predictability, respectively.

• B8 - Role of uncertainty in ice microphysical processes in warm conveyor belts
  (PIs: Grams, Christian; Hoose, Corinna; Miltenberger, Annette)

  In the warm sector of an extratropical cyclone, a rapidly ascending air stream - the warm conveyor belt - transports air from low levels to the tropopause. It modifies the flow in the upper troposphere and the evolution of downstream weather. However, the numerical description of clouds in the warm conveyor belt, which contain liquid water and ice, is still uncertain. Here we will investigate how the uncertainty of ice microphysical processes propagates to the warm conveyor belt outflow. For this, we will use nested simulations with ICON, creating a large ensemble and evaluate it by means of Lagrangian diagnostics and statistical emulation.

The WG “Clouds and storm tracks” is involved in one sub-project of “Waves to Weather”:

• B4 - Radiative interactions at the NWP scale and their impact on midlatitude cyclone predictability
  (PIs: Mayer, Bernhard; Voigt, Aiko)

  Diabatic heating and cooling by radiation is the main driver of atmospheric circulation. The project focuses on aspects of radiation that will become increasingly important as models move to higher and higher resolution. It will identify ways to calculate radiation more efficiently and more accurately, including 3-dimensional radiative effects, and it will study the dynamical impact through simulations of midlatitude cyclones. The twin-project combines the expertise of LMU and KIT in radiative transfer, high-resolution modeling with ICON, and cloud-radiation-circulation coupling.

The WG “Atmospheric Risks” is involved in one sub-project of “Waves to Weather”:

• B1 - Microphysical uncertainties in hailstorms using statistical emulation and stochastic cloud physics
  (PIs: Hoose, Corinna; Kunz, Michael; Miltenberger, Annette)

  Severe convective storms, especially if associated with large hail, frequently cause substantial damage. When and where hailstorms occur and how intense they are, is very difficult to forecast. In this project, we use advanced statistical methods to investigate to what extent the forecast quality and reliability can be improved by reducing uncertainties in the numerical description of clouds, taking into account the additional uncertainty in the knowledge of relevant environmental parameters such as temperature, humidity and aerosol concentration.

The WG “Regional Climate and Weather Hazards” is involved in one sub-project of “Waves to Weather”:

• C8 - Stratospheric influence on predictability of persistent weather patterns
  (PIs: Birner, Thomas; Garny, Hella; Pinto, Joaquim G.)

  Stratospheric circulation anomalies during Northern winter/spring may enhance the predictability of large-scale tropospheric circulation patterns that can lead to weather extremes in mid-latitudes (e.g. the recent March 2018 blocking event and cold spell over continental Europe). This project will use specifically designed re-forecast experiments to provide new insights into how stratospheric circulation anomalies may influence predictability of persistent tropospheric weather patterns, in particular intense and persistent blocking events (linked to cold spells), and extended storm series (gusts and precipitation).