Richard Müller

Floods occur frequently around the world and belong to the most dangerous meteorological hazards, causing destruction of infrastructure and loss of human life. Therefore, the estimation and nearcast of precipitation is of uttermost importance to launch warnings early enough.Although radars are a reliable source for precipitation estimation, their coverage is not sufficient for comprehensive monitoring and near real-time observation of precipitation in many regions, e.g. radar data are largely unavailable over the ocean, hence largely blind regarding thunderstorms coming from the sea. Also over complex terrain and in remote areas radar information is only sparely available. Further, saturation effects hamper the accurate prediction of rain rates of large thunderstorms and finally rivers or catchment areas do not end close to national borders, but national radar networks do in contrast to satellite data.Could satellite based precipitation data fill the gaps ? Up to now satellite based precipitation was not accurate enough to supplement radar data and to improve the spatial coverage and accuracy of precipitation rates around the world. Therefore, a novel method based on artificial intelligence has been developed to achieve an accuracy close to that of radar based rain rates. In this method a precipitation rate is derived every ten minutes from the satellite-based effective cloud albedo by regression with rain gauge data (ombrometer). This approach provides a first near real time Quantitative Precipitation Estimation (QPE) to calculate satellite-based precipitation rates with a large geographical coverage. However, this approach shows shortcomings in regional differentiation as the regression is applied over all existing cloud types. In a second step, a method is therefore applied in order to learn from cloud structures (gradients, curvatures) and in this manner to take into account regional differences in the relationship between the retrieved cloud information and the precipitation rates. The method automatically transfers the learned relationship to regions without rain gauge data, a process that the authors refer to as spatial transfer learning. Thus, a meaningful QPE is also possible in areas without any rain gauge data (ombrometer), e.g. over the ocean and countries without radar data. Of course the method can also be applied to radar data and could improve the QPE from radar as well. Hence, the method will be also applied to radar.The presentation will provide an overview about the developed method and the validation results with a focus on satellite based rain rates. However, the application to radar will be briefly discussed as well. The presentation will close with a brief outlook.

High ice water content (HIWC) conditions are a concern for aviation as the ingestion of ice particles in the jet engines can induce ice crystal icing (ICI), which results in performance loss and damage. To constantly monitor these conditions, retrievals for the detection of ICI were recently developed based on geostationary satellite imagery, but their calibration is limited to targeted flight campaigns or scattered samplings from ICI events databases. In this work, we close this gap, using exclusively remote sensing data to develop and assess a new retrieval for potential ICI conditions. Cloud IWC measurements are provided from the synergy of radar and lidar (DARDAR) on board the polar-orbiting satellites CloudSat and CALIPSO. HIWC conditions (IWC≥0.5 g m−3) at typical cruise altitudes are used as the proxy for areas with potential ICI formation. The HIWC conditions predictors are taken from a combination of observations and retrievals of the geostationary satellite Meteosat Second Generation (MSG). A random forest is trained and tested based on the collocated dataset of active and passive measurements during the summer months of 2013 and 2015, covering the European domain. The input predictors are the brightness temperature difference between the MSG channels at 6.2 and 10.8 µm wavelengths, the visible channel at 0.6 µm wavelength, the cloud optical thickness at 0.6 µm wavelength, and four convection metrics related to the distance to the closest convective cell, area extent of the convective cells, and convection density in the pixel surroundings. Over Europe, 83 % of HIWC conditions measured in the DARDAR dataset are correctly detected. The associated false alarm rate is 51 %. The retrieval is further tested with the ICI events database reported by Lufthansa. Four out of seven events are correctly detected. In conclusion, the retrieval achieves performances comparable to previously developed retrievals. An operational application would enable aircraft rerouting around areas with high ICI probability.

High ice water content (HIWC) conditions are a concern for aviation as the ingestion of ice particles in the jet engines can induce ice crystal icing (ICI), which results in performance loss and damage. To constantly monitor these conditions, retrievals for the detection of ICI were recently developed based on geostationary satellite imagery, but their calibration is limited to targeted flight campaigns or scattered samplings from ICI events databases. In this work, we close this gap, using exclusively remote sensing data to develop and assess a new retrieval for potential ICI conditions. Cloud IWC measurements are provided from the synergy of radar and lidar (DARDAR) on board the polar-orbiting satellites CloudSat and CALIPSO. HIWC conditions (IWC ≥ 0.5 g m⁻³) at typical cruise altitudes are used as the proxy for areas with potential ICI formation. The HIWC conditions predictors are taken from a combination of observations and retrievals of the geostationary satellite Meteosat Second Generation (MSG). A random forest is trained and tested based on the collocated dataset of active and passive measurements during the summer months of 2013 and 2015, covering the European domain. The input predictors are the brightness temperature difference between the MSG channels at 6.2 and 10.8 µm wavelengths, the visible channel at 0.6 µm wavelength, the cloud optical thickness at 0.6 µm wavelength, and four convection metrics related to the distance to the closest convective cell, area extent of the convective cells, and convection density in the pixel surroundings. Over Europe, 83 % of HIWC conditions measured in the DARDAR dataset are correctly detected. The associated false alarm rate is 51 %. The retrieval is further tested with the ICI events database reported by Lufthansa. Four out of seven events are correctly detected. In conclusion, the retrieval achieves performances comparable to previously developed retrievals. An operational application would enable aircraft rerouting around areas with high ICI probability.

High ice water content (HIWC) conditions are a concern for aviation as the ingestion of ice particles in the jet engines can induce ice crystal icing (ICI), which results in performance loss and damage. To constantly monitor these conditions, retrievals for the detection of ICI were recently developed based on geostationary satellite imagery, but their calibration is limited to targeted flight campaigns or scattered samplings from ICI events databases. In this work, we close this gap, using exclusively remote sensing data to develop and assess a new retrieval for potential ICI conditions. Cloud IWC measurements are provided from the synergy of radar and lidar (DARDAR) on board the polar-orbiting satellites CloudSat and CALIPSO. HIWC conditions (IWC ≥ 0.5 g m⁻³) at typical cruise altitudes are used as the proxy for areas with potential ICI formation. The HIWC conditions predictors are taken from a combination of observations and retrievals of the geostationary satellite Meteosat Second Generation (MSG). A random forest is trained and tested based on the collocated dataset of active and passive measurements during the summer months of 2013 and 2015, covering the European domain. The input predictors are the brightness temperature difference between the MSG channels at 6.2 and 10.8 µm wavelengths, the visible channel at 0.6 µm wavelength, the cloud optical thickness at 0.6 µm wavelength, and four convection metrics related to the distance to the closest convective cell, area extent of the convective cells, and convection density in the pixel surroundings. Over Europe, 83 % of HIWC conditions measured in the DARDAR dataset are correctly detected. The associated false alarm rate is 51 %. The retrieval is further tested with the ICI events database reported by Lufthansa. Four out of seven events are correctly detected. In conclusion, the retrieval achieves performances comparable to previously developed retrievals. An operational application would enable aircraft rerouting around areas with high ICI probability.

High ice water content (HIWC) conditions are a concern for aviation as the ingestion of ice particles in the jet engines can induce ice crystal icing (ICI), which results in performance loss and damage. To constantly monitor these conditions, retrievals for the detection of ICI were recently developed based on geostationary satellite imagery, but their calibration is limited to targeted flight campaigns or scattered samplings from ICI events databases. In this work, we close this gap, using exclusively remote sensing data to develop and assess a new retrieval for potential ICI conditions. Cloud IWC measurements are provided from the synergy of radar and lidar (DARDAR) on board the polar-orbiting satellites CloudSat and CALIPSO. HIWC conditions (IWC ≥ 0.5 g m⁻³) at typical cruise altitudes are used as the proxy for areas with potential ICI formation. The HIWC conditions predictors are taken from a combination of observations and retrievals of the geostationary satellite Meteosat Second Generation (MSG). A random forest is trained and tested based on the collocated dataset of active and passive measurements during the summer months of 2013 and 2015, covering the European domain. The input predictors are the brightness temperature difference between the MSG channels at 6.2 and 10.8 µm wavelengths, the visible channel at 0.6 µm wavelength, the cloud optical thickness at 0.6 µm wavelength, and four convection metrics related to the distance to the closest convective cell, area extent of the convective cells, and convection density in the pixel surroundings. Over Europe, 83 % of HIWC conditions measured in the DARDAR dataset are correctly detected. The associated false alarm rate is 51 %. The retrieval is further tested with the ICI events database reported by Lufthansa. Four out of seven events are correctly detected. In conclusion, the retrieval achieves performances comparable to previously developed retrievals. An operational application would enable aircraft rerouting around areas with high ICI probability.

High ice water content (HIWC) conditions are a concern for aviation as the ingestion of ice particles in the jet engines can induce ice crystal icing (ICI), which results in performance loss and damage. To constantly monitor these conditions, retrievals for the detection of ICI were recently developed based on geostationary satellite imagery, but their calibration is limited to targeted flight campaigns or scattered samplings from ICI events databases. In this work, we close this gap, using exclusively remote sensing data to develop and assess a new retrieval for potential ICI conditions. Cloud IWC measurements are provided from the synergy of radar and lidar (DARDAR) on board the polar-orbiting satellites CloudSat and CALIPSO. HIWC conditions (IWC ≥ 0.5 g m⁻³) at typical cruise altitudes are used as the proxy for areas with potential ICI formation. The HIWC conditions predictors are taken from a combination of observations and retrievals of the geostationary satellite Meteosat Second Generation (MSG). A random forest is trained and tested based on the collocated dataset of active and passive measurements during the summer months of 2013 and 2015, covering the European domain. The input predictors are the brightness temperature difference between the MSG channels at 6.2 and 10.8 µm wavelengths, the visible channel at 0.6 µm wavelength, the cloud optical thickness at 0.6 µm wavelength, and four convection metrics related to the distance to the closest convective cell, area extent of the convective cells, and convection density in the pixel surroundings. Over Europe, 83 % of HIWC conditions measured in the DARDAR dataset are correctly detected. The associated false alarm rate is 51 %. The retrieval is further tested with the ICI events database reported by Lufthansa. Four out of seven events are correctly detected. In conclusion, the retrieval achieves performances comparable to previously developed retrievals. An operational application would enable aircraft rerouting around areas with high ICI probability.

High ice water content (HIWC) conditions are a concern for aviation as the ingestion of ice particles in the jet engines can induce ice crystal icing (ICI), which results in performance loss and damage. To constantly monitor these conditions, retrievals for the detection of ICI were recently developed based on geostationary satellite imagery, but their calibration is limited to targeted flight campaigns or scattered samplings from ICI events databases. In this work, we close this gap, using exclusively remote sensing data to develop and assess a new retrieval for potential ICI conditions. Cloud IWC measurements are provided from the synergy of radar and lidar (DARDAR) on board the polar-orbiting satellites CloudSat and CALIPSO. HIWC conditions (IWC ≥ 0.5 g m⁻³) at typical cruise altitudes are used as the proxy for areas with potential ICI formation. The HIWC conditions predictors are taken from a combination of observations and retrievals of the geostationary satellite Meteosat Second Generation (MSG). A random forest is trained and tested based on the collocated dataset of active and passive measurements during the summer months of 2013 and 2015, covering the European domain. The input predictors are the brightness temperature difference between the MSG channels at 6.2 and 10.8 µm wavelengths, the visible channel at 0.6 µm wavelength, the cloud optical thickness at 0.6 µm wavelength, and four convection metrics related to the distance to the closest convective cell, area extent of the convective cells, and convection density in the pixel surroundings. Over Europe, 83 % of HIWC conditions measured in the DARDAR dataset are correctly detected. The associated false alarm rate is 51 %. The retrieval is further tested with the ICI events database reported by Lufthansa. Four out of seven events are correctly detected. In conclusion, the retrieval achieves performances comparable to previously developed retrievals. An operational application would enable aircraft rerouting around areas with high ICI probability.

Machine learning has a significant impact on severe weather nowcasting, with purely data-driven models quickly moving into the focus of current research over traditional physically motivated methods. Yet, recent work on radar-based precipitation nowcasting suggests that blending physical insight—like advection—with deep learning can improve forecast skill. In this talk, we will present our attempt to test how general this idea really is by applying it to satellite-based thunderstorm nowcasting for the first time.The central question we address is when and why advection should improve ML-based forecasts. Using a simple scale argument, we show that advection through optical flow algorithms helps preserve storm patterns within a model’s receptive field, which is critical for nowcasts at longer lead times. To test this, we trained convolutional neural networks to nowcast thunderstorms based on satellite imagery and lightning observations, with and without incorporating advected inputs.When considering average performance, the impact is modest. But once we break down the results by lead time and wind speed, a clear pattern emerges: the benefit of including advection emergesafter about 2 hours of lead time and grows with increasing wind speed, in agreement with our physical reasoning.Our findings highlight the importance of considering physical scales when designing and evaluating ML-based forecasting systems.

The ability to produce reliable short-range forecasts for thunderstorms is crucial for issuing timely emergency warnings to the general population, protecting vital infrastructure and alerting first responders in advance. Regularly affected by thunderstorms and associated hazards are remote areas, mountainous regions and air traffic, necessitating broadly available nowcasting solutions. Geostationary satellites provide an ideal data source for thunderstorm nowcasting in these cases.Traditional approaches employ deterministic algorithms to predict the occurrence and evolution of thunderstorms essentially by solving the optical flow problem. Recently, machine learning (ML) has emerged as an alternative and already shown promising results. While U-Net-based architectures have proven to be a very robust baseline, ML-based methods currently represent an extremely active area of research and optimal approaches have yet to crystallise. Particularly for satellite-based thunderstorm nowcasting, the superiority of either traditional methods or a specific ML architecture has yet to be established.It is our goal to narrow this knowledge gap by performing an extensive benchmark. In this talk we present a preliminary study featuring a set of nowcasting tools continuously improved over the past 20 years at the German Aerospace Centre (DLR) as well as a newly developed ML-model. The presented evaluation pipeline is developed jointly with the German Meteorological Service (DWD) and the regional Italian meteorological service Arpae Emilia-Romagna as part of the Italia-Deutschland Science-4-Service network. It serves as a baseline for a broader follow-up study including satellite-based thunderstorm nowcasting operationally used at DWD and Arpae. All models are evaluated over a period of multiple years for a fixed region in central Europe, using only satellite imagery from the SEVIRI instrument onboard MSG as input, while lightning recorded by the LINET network serves as ground truth. The accuracy of each model is measured for a set of lead times up to 180 minutes according to established metrics in a homogeneous setting.Our work aims to guide future research and development efforts, ultimately paving the way for improved satellite-based thunderstorm nowcasting.

Observations from weather radars are well established for the estimation of precipitation. However these data are unavailable over the ocean and only sparely available over remote regions. Furthermore, in mountains the retrieval of precipitation is severely compromised by the scan geometry. Finally, non-meteorological noise and the saturation of the radar signal by thunderstorm clouds also pose seroius problems to accuracy. Satellite data can help to overcome these shortcomings. However, the infrared signal provides mainly information about the cloud top temperature and thus no information about rain droplets. However, during day there is a good correlation of the cloud optical thickness and effective radii as e.g. demonstrated in the PhD of Rob Roebeling. Nevertheless in any case the satellite based precipitation needs to be calibrated. Typically, satellite instruments measuring in the micro-wave are used for this purpose. However, in this study, we use ground based data (omrbometer) for the calibration of the satellite based precipitation rate. We investigate various machine learning methods (e.g. SVM, transformer based GANs and transfer learning) to find the best way to perform recurrent, near-real time calibration updates based on pre-trained information. In this way, the spatial information of the satellite precipitation can be optimally combined with the high accuracy of the ground based measurements. The presentation will provide an overview about the developed methods and the validation results and the link of the work to the WMO-AINNP project. Further a outlook will be given about the nowcasting of satellite based precipitation and its integration into the concept of seamless prediction at DWD and the combination with radar nowcasting.