Global warming is leading to an increase in the frequency and intensity of extreme weather events. In arid and semi-arid regions such as the Tensift basin in Morocco, global warming reflected in severe water shortages. In this context, it is of crucial importance to precisely quantify the amount of precipitation and its variability with respect of time and space. In the absence of adequate coverage by weather stations, it is imperative to investigate the reliability of satellite-based precipitation products in estimating observed precipitation. The objective of this study is to assess the performance of eight satellite products, namely PERSIANN, PERSIANN CDR, IMERG, ARC2, RFE2, CHIRPS, ERA5, and MSWEP in accurately estimating extreme observed precipitation within the basin. To this end, satellite precipitation products were compared to observed data collected over fourteen weather stations between 2001-2016, at various time steps (daily, monthly, seasonal, and annual). The data were analyzed using volumetric and categorical metrics. We further evaluated the estimation of the satellite products for extreme precipitation by comparing their extreme indices with those of the weather stations. In the same vein, this study examined the behavior of satellite products during droughts by comparing their SPI with the observations to determine the most accurate satellite products in determining the onset, duration, and magnitude of droughts. Finally, in order to determine whether bias correction was necessary to enhance performance and identify the most effective method, in the context of extreme events, several bias correction methods were evaluated based on the same procedures before and after bias correction. The results showed that PERSIANN CDR, IMERG, MSWEP, and ERA5 are the most accurate, with better performance at both monthly and annual time steps relative to daily time steps. Seasonally, the worse performance of all products occurs in summer and the best in autumn. Furthermore, we found out that PERSIANN CDR is the most reliable product that water managers can use for extreme events, while MSWEP, ERA5, and PERSIANN CDR are the best products that can be used to study the impact of climate change on drought. Finally, The Cumulative Distribution Function (CDF) mapping was found to be the most effective method in correcting the bias in the Tensift basin, especially for extreme events.

How to cite: Salih, W., Epule Epule, T., El mahdi, E. K., and Chehbouni, A.: Assessment of Satellite Precipitation Products during Extreme Events in a Semiarid Region, EMS Annual Meeting 2023, Bratislava, Slovakia, 4–8 Sep 2023, EMS2023-122, https://doi.org/10.5194/ems2023-122, 2023.

Food and economic safety in West Africa rely heavily on rainfed agriculture and are threatened by climate change and demographic growth. Accurate rainfall information is therefore crucial to ensure safety against these challenges. However, existing rainfall models fail to accurately represent the highly variable and sparsely monitored West African rainfall distribution. Satellite rainfall products show a poor correlation with ground-based rainfall measurements and literature suggests that atmospheric aerosols are partly to blame for this poor performance.

To address this challenge, we propose a Deep Learning (DL) model that utilizes satellite water vapor (WV) and Thermal Infrared (TIR) observations in conjunction with temporal information for satellite rainfall retrieval in West Africa. We leverage the TIR and WV channels of the Meteosat Second Generation satellite to develop a DL model for satellite rainfall detection. Our results indicate that incorporating WV data into the DL framework enables the detection of strong convective motions typically associated with heavy rainfall. This is particularly relevant in regions where convective rainfall is dominant, such as the tropics.

Furthermore, the WV data facilitate the identification of dry air masses advected from the nearby Sahara Desert, which often create discontinuities in precipitation events over our study area. The ability to detect such dry air masses is a significant advantage of our proposed DL model, as it aids to reduce false alarms and rainfall overestimation as compared to methods that rely only on TIR data.

Our DL model achieves a robust performance in rainfall binary classification, with fewer false alarms and lower rainfall overdetection (FBias < 2.0) than the state-of-the-art Integrated Multi-satellitE Retrievals for GPM (IMERG) Final Run. Our findings suggest that incorporating WV observations and temporal information in a DL framework efficiently complement TIR observations and enhances the accuracy of satellite rainfall retrieval in West Africa.

How to cite: Estebanez Camarena, M., Curzi, F., Taormina, R., van de Giesen, N., and ten Veldhuis, M.-C.: RainRunner: A Deep Learning satellite rainfall retrieval  model for West Africa, EMS Annual Meeting 2023, Bratislava, Slovakia, 4–8 Sep 2023, EMS2023-551, https://doi.org/10.5194/ems2023-551, 2023.

Traditional tracking algorithms use a single threshold in a precipitation or infrared brightness temperature field to identify and track precipitation systems. Though valuable, these algorithms have limitations in tracking Mesoscale Convective Systems (MCSs) that sometimes occur as clusters embedded in synoptic scale disturbances such as tropical and mid-latitude waves. These embedded systems might be connected but should not be identified as a single large system since the connection is transient. The recent detect and spread (DAS) type algorithms help identify an MCS in a cluster  by identifying a core region of heavy rainfall and spreading it into adjacent regions of lower precipitation values. Also, as we move away from a single satellite and towards microsatellite constellations for various meteorological data, we need a robust method to identify and track objects of interest in a multi-satellite product. This is because each satellite in the constellation may have a different sensor that requires cross-calibration and is finally merged to create a single product.

 We present the improved version of the Forward in Time (FIT) tracking program, a multi-threshold, detect and spread type algorithm, to track MCSs in Integrated MultisatellitE Retrievals for Global Precipitation mission (IMERG), NASA's global precipitation product. IMERG is a  multi-satellite precipitation product that combines rain retrievals from passive microwave sensors on a virtual constellation of satellites and rain retrievals from Infrared sensors onboard geostationary satellites. Using the FiT algorithm, we track MCSs in the IMERG precipitation field for ten years (2011-2020) and store MCSs' properties in a publicly available dataset called Tracked IMERG Mesoscale Precipitation Sytems (TIMPS). Leveraging this dataset, we present the regional variability of MCSs properties (frequency, lifetime, and propagation velocity) and some preliminary results from ongoing studies.

How to cite: Rajagopal, M., Skok, G., Russell, J., and Zipser, E.: A multi-threshold-based identification and tracking of Mesoscale Convective Systems  in a multi-satellite precipitation product, EMS Annual Meeting 2023, Bratislava, Slovakia, 4–8 Sep 2023, EMS2023-396, https://doi.org/10.5194/ems2023-396, 2023.

The hydrological cycle in southern France is influenced by highly variable precipitation, which impacts the variability of soil moisture and continental runoff. The regional hydrological cycle is also significantly impacted by the Mediterranean Sea (Gulf of Lion) through moisture advection, and the strong coupling between the atmosphere and land surface affects all branches of the regional hydrological cycle, particularly due to the region's special orographic structure. The potential intensification of the regional hydrological cycle under climate change scenarios, leading to extreme hydroclimate events, is a critical factor for understanding the mechanisms of climate variability in the Mediterranean region.

In this study, long-term daily precipitation data from 1979 to 2014 for southern France was analyzed using various datasets, including data from 300 stations in the METEO-FRANCE collection, modern-era reanalysis (ERA5, JRA55), and satellite datasets. Diagnostic analysis was conducted to assess linear trends and interannual variability in total precipitation and precipitation extremes using multiple datasets. The results of this analysis allowed for the quantification of differences in variability patterns of precipitation among the datasets, highlighting their respective strengths and weaknesses. The study also examined the regional responses of soil moisture to precipitation using data from the GLEAM-GLDAS datasets besides 14 SMOSANIA stations. Finally, the representation of the observed regional hydrological cycle in historical simulations with regional (Euro-CORDEX) and global (CMIP6) climate models are analyzed. Thus, a statistical association between precipitation, soil moisture, and continental runoff is established and the role of regional atmospheric circulation and atmospheric rivers in forming multidecadal changes in the regional hydrological cycle is quantified.

How to cite: Jomaa, F. and Zolina, O.: Precipitation over coastal regions of southern France and its impacts on the regional hydrological cycle, EMS Annual Meeting 2023, Bratislava, Slovakia, 4–8 Sep 2023, EMS2023-500, https://doi.org/10.5194/ems2023-500, 2023.

Over the last century, climate change has become a subject of great interest and is recognized as one of the major global challenges facing humankind in the 21st century. Changing precipitation patterns including extreme events such as more intense and prolonged dry periods have become a growing concern for people living in the Pacific Island region given their high dependence on rainfall for their freshwater needs as well as the precipitation impact on different sectors and socioeconomic activities. However, up to date, little attention has been paid to understanding the implications of climatic changes for people and their capacity to manage these changes. Here we have analyzed the temporal evolution of precipitation data over the Fiji islands for the period 1905–2021, using observational in-situ data from 23 meteorological stations and using 23 climate indices calculated with the ClimInd R package by using daily precipitation data. These indices are focused on extreme precipitation events which characterizes the data record as regards the frequency (e.g. the number of dry days or the number of very wet days), the duration (e.g. the longest dry period or the longest wet period) and the precipitation intensity (e.g. the total precipitation amount) among others. Positive increasing trends were found for the majority of daily climate indices although no statistically significant trends were dominant. The results derived from this study can be useful as validation of climate models helping to better understand climate change processes in this vulnerable climate region from a wide perspective which could help for climate decision-making.

 Keywords: precipitation pattern, extreme events, Pacific Island region, trends, climate indices, ClimInd.

How to cite: Fernández-Duque, B., Vicente-Serrano, S. M., Kumar, R., Domínguez-Castro, F., Peña-Angulo, D., Noguera, I., and El Kenawy, A.: Assessment of daily and monthly precipitation record data over the Fiji Islands for the period 1905-2021, EMS Annual Meeting 2023, Bratislava, Slovakia, 4–8 Sep 2023, EMS2023-128, https://doi.org/10.5194/ems2023-128, 2023.

For over 20 years, the Operational Programme on the Exchange of Weather Radar Information (OPERA) has been coordinating radar cooperation among national weather services in Europe, operating under the umbrella of the European Meteorological Services Network, EIG EUMETNET. More information on this programme can be found at www.eumetnet.eu/opera [Saltikoff et al. 2019].

Since 2011, the OPERA data center (ODC), also known as Odyssey, has been producing three Pan-European radar composite products, with spatial resolution of 2 km and 15-minute update cycle. The composite products are:  maximum reflectivity, rain rate, and 1-hour accumulation. Additionally, Odyssey provides the quality-controlled (QC) single-site volume radar data for the numerical weather prediction (NWP) consortia to be used for assimilation. OPERA data are utilized for a variety of applications, including nowcasting, NWP assimilation, and flood warnings. As the OPERA data archive has grown to over ten years, also the creation of climatological applications is possible based on OPERA data. Different user groups have various needs for the data, with some requiring advanced QC and production of complex products, while others prioritize timely access to the data.

During the current OPERA phase 5 (2019-2023), the programme's main focus is to gradually replace the ODC with three new production lines:

• The CUMULUS/STRATUS line is responsible for gathering incoming radar data and forwarding it to other production lines.
• The CIRRUS line generates a high-resolution maximum reflectivity composite (updated every 5 minutes with 1 km horizontal resolution).
• The NIMBUS line produces precipitation composites and provides quality-controlled volume radar data for the purpose of NWP assimilation.

Additionally, OPERA centrally monitors the network's performance, including the level of radio frequency interference and the antenna pointing using methods such as solar monitoring [Huuskonen and Holleman, 2007].

During the conference, we will present the updated production lines and preliminary results of their performance compared to the ODC. The CIRRUS represents an updated version of the ODC software, but due to the different spatial and temporal resolutions, the CIRRUS maximum reflectivity product is inherently different from the ODC equivalent composite. The methodology used to compare these two products is based on the SAL method [Wernli et al. 2008], which not only compares the amplitude but also the structural differences of the two products. The NIMBUS production is based on the BALTRAD open radar software, and its performance for the rain accumulation composite product is compared to gauge observations, similar as in the study by [Park et al. 2019].

We will also provide insights for data sharing in OPERA. The EU Implementing Regulation has defined weather radar data as High-Value Datasets (HVD) that should be shared free of charge, openly accessible via Application Programming Interfaces (APIs), machine-readable, and bulk downloadable in the coming years.

How to cite: von Lerber, A., Goreta, M., Haase, G., Jurašek, M., Karsisto, P., Klink, S., Koutek, M., Leijnse, H., Meyer, V., Müller, C., Nicolau, T., Park, S., Peura, M., Radojevic, M., Stephan, K., Tüchler, L., and Vodarić Šurija, B.: OPERA5 – the renewal of the production lines, EMS Annual Meeting 2023, Bratislava, Slovakia, 4–8 Sep 2023, EMS2023-325, https://doi.org/10.5194/ems2023-325, 2023.

We present the results of a feasibility study on the possibility of using the OPERA data information model (ODIM) with HDF5 file format implementation [Michelson et. al. 2021] for reprocessing long-term weather radar data series in the context of a small national meteorological service such as the Slovak Hydrometeorological Institute (SHMU). 

ODIM was developed during the Operational Programme on the Exchange of Weather Radar Information (OPERA) under the umbrella of the European Meteorological Services Network, EIG EUMETNET [Saltikoff et al. 2019] to support the exchange of volume radar data between meteorological services and different radar systems.

Work was performed in 3 consequent steps:

1. Radar archive consolidation

SHMU has been working with radar data since 1965 [Podhorský & Guba 2014], but the level of equipment available didn't allow us to digitally archive such a large amount of data. We were able to extract archived radar from 1998 to the present. Data from earlier periods were archived on analogue media (e.g. paper drawings, photographs) or on special tapes that we cannot read today. More than 50 old data tapes and more than 500 CDs and DVDs have been extracted. The most recent period was stored in the robotic tape library. We will discuss all the problems we encountered during this step.

2. Conversion to ODIM hdf5 files

From the archive we extracted data from 4 different radar systems: MRL-5, EEC DWSR92C, Radtec RDR-250GCDP with Gamic signal processor and Selex Meteor 735 CDP. 5 different converters were prepared. The most interesting was the converter for MRL-5 data. It only stores raw data, radar equation has to be applied, noise figure subtracted. The scan was not clearly defined, volumes store all data from start of measurement to end, sampled in any position of the antenna, and we got uncorrected reflectivity for S- and X-band channel, because MRL-5 was classic non-Doppler radar and clutter has to be filtered in processing.

3. Processing

For processing, we used "home-made" qRad software running on hpc. It is capable of generating any standard product and making a common composite based on 11+ different quality indices. The clutter map was calculated from the previous 2-3 week period. We had to solve problems with temporal and spatial variability of the input data. The scan repetition time had values of 30, 15, 7.5, 10 and 5 min time. The number of radars varied between 0 and 4. We decided to use only those dates when at least 2 radars were available and covered the whole of Slovakia. As a reference product, the composite of CAPPI 3km was chosen, which provides uniform data for all radars over whole Slovakia. Maps of various statistical variables were calculated from CAPPI 3km composites.

How to cite: Méri, L., Jurašek, M., Okon, Ľ., and Kaňák, J.: Long time radar data series processing using OPERA ODIM, EMS Annual Meeting 2023, Bratislava, Slovakia, 4–8 Sep 2023, EMS2023-337, https://doi.org/10.5194/ems2023-337, 2023.

Deep-Learning (DL) is a sub-field of Machine-Learning (ML) whose popularity has grown exponentially during the last decade thanks to its numerous successes related to artificial intelligence, e.g. visual object recognition and detection, image generation, natural language processing or speech recognition. But, like other ML algorithms, DL can be applied to a broader variety of problems, as long as the necessary data is available to train the model. In this aspect, the field of meteorology satisfy the data requirement. Particularly, DL could improve the accuracy of precipitation estimation. Accurate precipitation estimation is a very important product of weather institutes for water supply monitoring, and as input and validation for Numerical Weather Prediction models and precipitation nowcasting. Currently, precipitation are mainly estimated using both radar and rain-gauge data, but an important limitation of this method is the limited coverage of these measurements. A possible solution to this problem could come from satellite radiometers observations. Unfortunately, estimating precipitation accurately from satellite radiometer data is challenging.

As a solution, we developed a DL method to merge rain gauge measurements with a ground-based radar composite and satellite radiometer imagery. The proposed convolutional neural network, composed of an encoder–decoder architecture, performs a multiscale analysis of the three input modalities to estimate simultaneously the rainfall probability and the precipitation rate. We used SEVIRI infrared channels, the OPERA radar composite and the measurements of automatic rain gauges. The training of our model and its performance evaluation are carried out on a dataset spanning 5 years from 2015 to 2019 and covering Belgium, the Netherlands, Germany and the North Sea. Our results for instantaneous precipitation detection, instantaneous precipitation rate estimation, and for daily rainfall accumulation estimation show that the best accuracy is obtained for the model combining all three modalities. We show that the combination of rain gauge measurements with radar data allows for a considerable increase in the accuracy of the precipitation estimation, and the addition of satellite imagery provides precipitation estimates where rain gauge and radar coverage are lacking. We also show that our multi-modal model significantly improves performance compared to the European radar composite product provided by OPERA and the quasi gauge-adjusted radar product RADOLAN provided by the DWD for precipitation rate estimation. Additionally, we carried a study of our method on the case of the extreme precipitation event of July 2021 that affected Belgium and Germany, which caused huge societal and economical damage.

How to cite: Moraux, A., Dewitte, S., and Munteanu, A.: A deep learning multimodal method for precipitation estimation, EMS Annual Meeting 2023, Bratislava, Slovakia, 4–8 Sep 2023, EMS2023-567, https://doi.org/10.5194/ems2023-567, 2023.

The Ruisdael Observatory (https://ruisdael-observatory.nl/) is a network of advanced sensors and high-resolution models aimed at improving the accuracy of climate, weather, and air quality models at the regional scale in the Netherlands. Within Ruisdael, TU Delft oversees a broad network of over 30 sensors, ranging from basic weather stations to cutting-edge disdrometers, radiometers, micro-rain radars, X-band radars, and cloud radars.

In this presentation, we focus on analyzing the statistical and dynamic properties of raindrop size distributions recorded by our network of seven co-located optical disdrometers and micro-rain radars in the Rotterdam-Delft area. Our analysis covers a range of precipitation intensities, from drizzle to heavy convective rain exceeding 140 mm/h. We document some of the most intriguing cases we've encountered during the first three years of operation and perform case studies to understand the temporal variability of raindrop sizes, their moments, scaling properties, and vertical evolution from the melting layer down to the ground. We also investigate whether temperature has any notable influence on the shapes, characteristic sizes, and concentrations of the raindrop size distributions, which is important for understanding future rainfall extremes. 

Finally, we discuss some of our future plans for the observatory, including efforts to improve high-resolution precipitation measurements in the Netherlands. By providing a better understanding of the statistical and dynamic properties of raindrop size distributions, we can enhance our ability to forecast precipitation, mitigate the impacts of heavy rain events, and improve air quality predictions. The Ruisdael Observatory is a crucial component of these efforts and represents a significant step forward in advancing our understanding of atmospheric processes in the Netherlands.

How to cite: Schleiss, M., Mackenzie, R., Castro Gonçalves, A., and Unal, C.: From drizzle to downpour: investigating the statistical and dynamical properties of raindrop size distributions in the Netherlands using a network of co-located disdrometers and micro-rain radars, EMS Annual Meeting 2023, Bratislava, Slovakia, 4–8 Sep 2023, EMS2023-419, https://doi.org/10.5194/ems2023-419, 2023.