ECSS2023-75
https://doi.org/10.5194/ecss2023-75
11th European Conference on Severe Storms
© Author(s) 2023. This work is distributed under
the Creative Commons Attribution 4.0 License.

Enhance object-based thunderstorm nowcasting with the use of next-generation satellite data at DWD

Cornelia Strube, Richard Müller, and Manuel Werner
Cornelia Strube et al.
  • Deutscher Wetterdienst, Frankfurter Str 135, Offenbach, Germany

The first satellite of the European next-generation meteorological satellite series Meteosat Third Generation was launched in December 2022 and is expected to deliver data operationally by the end of 2023. The new Flexible Combined Imager (FCI) instrument provides an image repeat rate of 10 minutes and higher spatial resolution of the up to 500m at the sub-satellite point. In addition, new datasets from channels that have not been available from the predecessor instrument SEVIRI on Meteosat Second Generation as well as the new Lightning Imager instrument will offer improved opportunities for the thunderstorm nowcasting over Europe and Africa.

At Deutscher Wetterdienst (DWD; english: German Meteorological Service), we strive to include these new datasets into our nowcasting pipelines. Here, we want to present work on the definition and incorporation of satellite-based storm features to enhance the object-based thunderstorm nowcasting tool KONRAD3D. So far, satellite data was not part of the process. We define storm objects in satellite images using a region-growing approach seeded on the radar-based KONRAD3D feature centroids in particular in early-convection environments. These storm objects are then used to calculate properties of the developing storm from satellite data, like relevant convective (initiation) characteristics, e.g. glaciation, updraft, and cloud-top cooling properties from infrared brightness temperatures, as well as structural features in the visible reflectance. By design, these additional storm characteristics from the satellite can be easily assigned to the corresponding KONRAD3D object and used in decisions like the determination of the storm severity. In addition, we plan to use the statistical analysis of the properties and tracking of the radar and satellite objects alongside each other to develop a data-driven classification model to connect early detection of cumulus clouds from the satellite to the later severity of the developed storm. Since the commissioning phase data from FCI will only become available in the course of this year, the development still builds on equivalent channels in SEVIRI data for now prepared to switch to FCI as soon as the data becomes available.

How to cite: Strube, C., Müller, R., and Werner, M.: Enhance object-based thunderstorm nowcasting with the use of next-generation satellite data at DWD, 11th European Conference on Severe Storms, Bucharest, Romania, 8–12 May 2023, ECSS2023-75, https://doi.org/10.5194/ecss2023-75, 2023.