The joint U.S.-Japan Global Precipitation Measurement (GPM) mission has passed a decade of operations, and continues to pursue research, dataset production, and outreach related to precipitation.  One key activity currently in development is the release of an improved “Version 08” of all GPM precipitation and latent heating products.

This presentation summarizes key improvements to the GPM products for which NASA has lead responsibility as we approach the release of Version 08.  For example, the Goddard Profiling (GPROF) algorithm has implemented a Machine Learning-based algorithm that shows solid improvements in computing retrievals from the constellation of partner satellite passive microwave sensors when compared to the GPM Microwave Imager (GMI) retrievals.  And all of the PMW retrievals, including from GMI, show improved validation scores.  The Combined Radar Radiometer Algorithm (CORRA) now incorporates additional emphasis on GMI (and Tropical Rainfall Measuring Mission [TRMM] Microwave Imager [TMI]) data in regions where the radar lacks skill.  This is principally the light precipitation and snow at high latitudes and over Central Asia in the winter.  Each algorithm is being adjusted to ensure continuity for each product across the boundary in 2014 between the TRMM and the GPM eras, as well as across the TRMM and GPM Core Observatory orbit boosts.  The U.S. Science Team’s Integrated Multi-satellitE Retrievals for GPM (IMERG) was upgraded to give more flexibility in using different-quality PMW sensors, and a new ML-based retrieval has been

The presentation also considers major issues that require continued attention, including the operational challenge of swarms of “small”, perhaps short-lived satellites, and planning for the next-generation multi-satellite product.

How to cite: Huffman, G., Kummerow, C., Olson, W., and Stocker, E.: Status and Development of Version 08 in the NASA GPM Activities, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-14297, https://doi.org/10.5194/egusphere-egu26-14297, 2026.

Under global climate change, extreme precipitation events are becoming more frequent and intense, posing increasing risks to hydrological and drought-related disasters. Hunan Province in southern China is particularly vulnerable, yet the long-term spatiotemporal evolution of extreme precipitation remains insufficiently understood. This study investigates the characteristics and driving features of extreme precipitation in Hunan Province to support disaster prevention and water resource management. Daily precipitation records from 97 national meteorological stations spanning 1961–2024 were analyzed using a suite of extreme precipitation indices defined by the WMO Expert Team on Climate Change Detection and Indices (ETCCDI). Heavy rainfall was characterized using the R50 threshold based on regional precipitation classification standards. Empirical Orthogonal Function (EOF) analysis, the Mann–Kendall trend and abrupt change tests, and Morlet wavelet analysis were applied to examine spatial patterns, temporal variability, abrupt shifts, and periodic signals. The results indicate an overall drying tendency in extreme precipitation across Hunan Province. Consecutive dry days (CDD) show a significant increasing trend, while consecutive wet days (CWD) decrease significantly. Although 75.3% of stations exhibit declining annual total precipitation (PRCPTOT), 66% show increasing extreme heavy precipitation (R99P), suggesting reduced mean precipitation but intensified extremes. Spatially, extreme precipitation exhibits a hierarchical structure consisting of large-scale regional coherence, topography-modulated counter-phase patterns, and localized fragmented distributions. Abrupt changes are concentrated mainly before the 1980s, particularly during the 1960s, with fewer change points detected after 1990, primarily in central Hunan. Significant periodicities are identified at 2.07–2.25 years, ~31 years, and ~60.3 years, corresponding to ENSO-related short-term variability, medium-to-long-term oscillations, and AMO-related ultra-long-term signals, respectively. Overall, extreme precipitation in Hunan Province is characterized by increasing aridity, heightened local extreme rainfall risks, and multi-scale climate modulation. These findings advance scientific understanding of extreme precipitation evolution in complex terrain and provide critical insights for improving regional forecasting and early warning systems. The contrasting trends—increasing drought risk alongside intensified extreme rainfall—highlight the urgent need for integrated adaptation strategies that enhance water resource management resilience and infrastructure preparedness under climate change.

How to cite: Li, Q. and Liu, L.: Spatiotemporal distribution and variation characteristics of extreme precipitation of Hunan Province, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-2390, https://doi.org/10.5194/egusphere-egu26-2390, 2026.

This contribution analyses temporal changes in thermodynamic variables, namely temperature, geopotential heights of pressure levels, specific cloud ice water content, specific cloud liquid water content, specific humidity, specific rain water content, and specific snow water content, as a function of ground temperature. The analysis is based on ERA5 reanalysis data extracted at the geographic coordinates 50.0° N, 15.0° E for the period 1940–2024, covering pressure levels from 1000 to 300 hPa and synoptic times at 00, 06, 12, and 18 UTC. The selected location corresponds to the city of Prague, Czech Republic. The objective of the study is to investigate whether the vertical profiles of the listed variables exhibit temporal changes and how these changes relate to increasing ground temperature. Particular attention is given to variations in gaseous humidity, liquid water, and solid-phase water within the vertical atmospheric column, and to their dependence on season and time of day.

How to cite: Sokol, Z. and Řeyáčová, D.: ERA5 Analysis (1940-2024) of Vertical Thermodynamic and Moisture Variability over Prague (1940-2024), EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-3427, https://doi.org/10.5194/egusphere-egu26-3427, 2026.

The frequency and intensity of precipitation have changed significantly in China as previously reported. A relevant behavior is the variability of precipitation, which describes temporal fluctuations of precipitation events. Yet it remains unclear how precipitation variability has changed at different timescales over China. In this study, we show that precipitation variability has increased significantly since the 1960s, averaging 2.3 % per decade across China. The increase exists across the synoptic to intraseasonal timescales. The increase in precipitation variability is evident in all seasons with the greatest rate in winter in percentage, which is approximately three times as much as that in summer. Regionally, precipitation variability has risen significantly in northwestern, northeastern, and southeastern China, but has decreased insignificantly along the wet-dry transition belt extending from the north to southwestern China. Compared to trends in mean and extreme precipitation, the increase of precipitation variability is more widespread and with greater magnitudes. The changes in the top 10 % extreme precipitation events contribute ∼75 % of the amplification of precipitation variability nationwide. In addition to long-term trend, summer precipitation variability over eastern China is modulated by the Pacific Decadal Oscillation. This study revealed robust increases in precipitation variability over China since the 1960s across different timescales, seasons, and regions, which have far-reaching impacts on droughts, floods, and water resource management.

How to cite: Mo, X., Zhang, W., and Zhou, T.: Increased Precipitation Variability at Multi-timescales in China since the 1960s, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-4653, https://doi.org/10.5194/egusphere-egu26-4653, 2026.

The Tibetan Plateau (TP) is home to the largest number of high-altitude lakes on Earth. Nam Co is the third largest lake on TP. Obtaining accurate precipitation data at the Nam Co basin scale is crucial for a deeper understanding of the water cycle and related atmospheric processes in cold high-altitude lake basins, and it also provides solid data support for the innovation of precipitation remote sensing inversion algorithms and the improvement of regional climate models. The Institute of Tibetan Plateau Research, Chinese Academy of Sciences, has established a multi-scale precipitation observation platform in the Nam Co basin, with the goal of accurately obtain precipitation data with high spatial and temporal resolution at the basin scale. The platform is equipped with an X-band dual-polarization weather radar, a micro-rain radar, a Double Fence Intercomparison Reference (DFIR) gauge, two raindrop spectrometers, and 24 rain gauges (including 5 T-200B weighing-type rain gauges and 19 Hobo tipping-bucket rain gauges) distributed around the lake area. The X-band dual-polarization weather radar is capable of monitoring the reflectivity and polarization characteristics of precipitation particles within the basin, while other instruments assist the radar in accurately estimating the amount of precipitation at the basin scale.

Uncertainty and bias in precipitation measurement significantly impact the accurate estimation of precipitation in cold high-altitude regions, making the correction of precipitation measurements from different rain gauges essential. The DFIR system, due to its high-precision observation capabilities, is used as the benchmark for precipitation measurement to correct the observations from T-200B and Hobo gauges. Quality control of radar data is crucial for achieving accurate quantitative precipitation estimation (QPE). Therefore, it is necessary to improve the quality of radar data through steps such as denoising, elimination of non-meteorological echoes, systematic error correction, bright band correction, and attenuation correction. Based on quality-controlled radar data, corrected rain gauge data, and raindrop spectrometer data, we have developed QPE methods to achieve accurate estimation of precipitation for the Nam Co basin.

How to cite: Chen, Y., Han, R., Wang, H., and Chen, M.: Quantitative precipitation estimation in the Nam Co Basin with X-band dual-polarization weather radar, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-4763, https://doi.org/10.5194/egusphere-egu26-4763, 2026.

Based on hourly precipitation data from Chinese automatic weather stations (2000-2020), this study employs the REOF method to classify plateau precipitation into three regions: Central and Eastern Tibet, Central Qinghai, and Northwestern Yunnan (Region I); Western Tibet, Western Qinghai, and Northern Gansu (Region II); and Southeastern Qinghai and the Western Sichuan Plateau (Region III). Plateau precipitation generally decreases from east to west, with Region I exhibiting the earliest peak and Region III the latest. Hourly extreme precipitation amounts and frequencies both show an "east-high, west-low" pattern, and frequencies of heavy rainfall is higher at the southeastern plateau, the rain intensity is stronger at the northeastern plateau. The peak time of hourly heavy precipitation in Region III is the earliest. Under the Bay of Bengal (BOB) storm influence, heavy precipitation concentrates in the northeastern and southern plateau, and the high-frequency zone is located in the southeastern plateau. The precipitation peaks occur from afternoon to night, and the peak time in Region III is the earliest, showing a distinct east-to-west delay. The peak time of the average rain intensity is earlier than the peak time of the number of the stations exceeding four precipitation thresholds, which means that extreme heavy rainfall occurs more frequently in the afternoon, while widespread heavy precipitation favors night. In Region III, the frequency peaks exceeding four precipitation thresholds occurs around 2100-2200 LST, indicating heavy rainfall induced by the BOB storm favors night in this area. The maximum contribution rates of the BOB storm-related hourly heavy precipitation are distributed over the eastern and southern plateau, with the 90th percentile precipitation contribution rate exceeding 60%, highlighting the prominence of short-duration heavy rainfall. Complex topography in the eastern and southern plateau characterized by valleys and intersecting terrain enhances convergence and uplift of warm, moist airflows from northward-moving BOB storms, further facilitating heavy precipitation generation.

How to cite: Luo, Q. and Chen, Y.: Analysis of Multi-Scale Characteristics of Plateau Precipitation under the Influence of Bay of Bengal Storms, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-7031, https://doi.org/10.5194/egusphere-egu26-7031, 2026.

Accurate rainfall observations are important for hydrological applications. However, rainfall exhibits strong spatial and temporal variability, resulting in significant uncertainties in areal rainfall products. Estimates of this spatial and temporal variability are needed for spatial interpolation and merging of rainfall products. Traditional rain gauge networks are often too sparse to resolve this variability. In this study, we make use of a unique high-density rain gauge network with a high temporal resolution (i.e. 5-min) over a three-year period (2022-2024) to quantify the spatial variability of rainfall over the whole of the Netherlands (about 1 gauge per 10km²). We investigated the spatial variability of rainfall at different temporal aggregation intervals by fitting climatological spherical semi-variograms, revealing a strong seasonal pattern. In addition, we examined the spatial dependency of rainfall in different directions to characterize anisotropy. Furthermore, this high-density network enables us to assess uncertainties in rainfall estimates across different spatial scales, ranging from weather radar pixels (~1 km²), satellite footprints (~10-100 km²) to catchments scales.

How to cite: Rombeek, N., Brauer, C., Hrachowitz, M., and Uijlenhoet, R.: Quantifying spatial rainfall variability using a country-wide high-density rain gauge network, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-9968, https://doi.org/10.5194/egusphere-egu26-9968, 2026.

Precipitation estimates over the Estonian radar composite domain change with accumulation length, season, and spatial location, yet these dependencies are rarely summarized in a compact way. This matters for comparing products across temporal resolutions and designing downstream applications such as precipitation modelling and nowcasting. Here we characterize precipitation structure using a Poisson-Gamma framework, focusing on the Tweedie variance-mean scaling exponent p as a descriptor of how wet-dry transitions and event-to-event fluctuations vary with aggregation.

Using 4-year radar composites aggregated from sub-hour to daily windows, we estimate p across accumulation lengths and derive seasonal maps that highlight where and when precipitation structure differs. In year-average, p increases with accumulation length across the domain, indicating that longer windows increasingly reflect the combined effect of multiple precipitation episodes within a window, while spatial gradients remain weak compared to the domain-wide shift with aggregation. Seasonal estimates show a consistent ordering, with p highest in summer and lowest in winter, and winter showing a stronger positive dependence on accumulation length. Seasonal maps at 6-24 h reveal clearer organization, including land-sea contrasts and enhanced spatial heterogeneity in warm and autumn seasons, whereas winter fields are smoother with localized marine features. We also compare radar-based behaviour with rain-gauge series at 10-min and 1-h temporal resolutions. Across common accumulation periods, p follows a consistent ordering, with higher values from the higher-temporal-resolution radar composite and lower values from the coarser gauge series, suggesting that temporal scaling influences inferred precipitation structure.

Overall, the study provides a set of figures and maps that summarize how precipitation structure varies with season and accumulation length over the Estonian radar domain. These results offer a baseline for multi-source comparison and for applications where aggregation scale and observing system matter, such as precipitation modelling, verification, and nowcasting-related target design.

How to cite: Tsoi, Y. C., Männik, A., and Rikka, S.: Scale- and Seasonal-dependent Precipitation Structure from Estonian Radar Composites Using a Poisson-Gamma Model, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-10353, https://doi.org/10.5194/egusphere-egu26-10353, 2026.

High-resolution, temporally consistent climate datasets are essential for hydrological modeling, water resource management, and climate impact assessments. The Po River District is the largest in Italy, spanning from the Alps to the plains, and exhibits substantial spatial heterogeneity in precipitation and temperature. However, existing datasets lack the spatial resolution necessary to capture the basin's diverse microclimates and complex orographic patterns, limiting their utility for process-based hydrological modeling and local-scale climate impact studies.

In this study, we generated a high-resolution (1 km x 1 km) daily gridded precipitation and temperature dataset over the Po River District. Following WMO standards, this 30-year (1991–2020) dataset provides a robust baseline for a region identified as one of Europe's most vulnerable climate change hotspots. The datasets were generated using the Kriging module available within the GEOframe-NewAGE modeling system, applied to quality-controlled ground station data. To address the vast area and topographic complexity, we implemented a spatial regionalization framework using Gaussian Mixture Models (GMM) to identify homogeneous climate zones. Zone-specific variogram models were derived and applied within the optimized Kriging framework. 

The model performance was rigorously evaluated using Leave-One-Out Cross-Validation (LOOCV) method. The validation results show exceptional accuracy for both variables. For temperature, the Kling-Gupta Efficiency (KGE) exceeded 0.75 at 99.7% of the stations, with strong correlations (>0.95). Notably for precipitation, over 80% of stations achieved KGE and correlation values above 0.75. The KGE decomposition revealed that errors primarily stemmed from variability estimation rather than bias, with 93% of stations showing optimal variance ratios (α = 0.75–1.25) and 99% maintaining near-unity bias (β ≈ 1).

This high-resolution dataset represents a significant advancement in regional climate data for the Po River District. The GMM-based regionalization successfully captured the basin's complex climatic regimes, enabling accurate spatial interpolation across diverse topographies. Beyond providing a WMO-compliant climatological baseline, these datasets are specifically designed to serve as high-resolution meteorological forcing input for distributed hydrological models, enabling process-based watershed simulations at unprecedented spatial detail. Future work will focus on coupling these datasets with the GEOframe-NewAGE hydrological modeling framework to assess the added value of 1-km climate forcing in capturing sub-basin scale hydrological responses, extreme event dynamics, and water balance components across the heterogeneous Po River landscape.

Acknowledgement

HS, JMW and RR would like to thank and acknowledge the funding support from Project “SPACE IT UP! ASI Contract n.2024-5-E.0 CUP Master n. I53D24000060005” SAP fund n: 000040104905.

How to cite: Salehi, H., Andreis, D., Roati, G., Wani, J. M., Brian, M., Tornatore, F., Formetta, G., and Rigon, R.: A 1-km Daily Gridded Climate Dataset for the Po River District (1991–2020): Regionalized Kriging within the GEOframe-NewAGE Framework, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-18192, https://doi.org/10.5194/egusphere-egu26-18192, 2026.

Abstract

Floods driven by short-duration intense rainfall, remain among the most damaging natural hazards in the Mediterranean and set major challenges for early warning systems. Accurate nowcasting (short-term forecasting) of convective rainfall is essential for hydrological response modelling and risk management. However, numerical weather prediction often struggles to capture storm initiation and localization in complex terrain.

This study investigates the assimilation of XPOL polarimetric radar data into the Weather Research and Forecasting (WRF) model using a 4DVAR data assimilation approach, to improve rainfall prediction for flood-relevant time scales. Selected high-impact precipitation events from 2024–2025 over Greece are simulated, including cases associated with flash flooding. Radar reflectivity and radial wind observations are assimilated through 4DVAR cycling, and simulations were performed at 2-km resolution with a 3-hour forecast horizon, representative of nowcasting. In addition, humidity, vertical velocity and horizontal wind divergence profiles estimated from lightning data, are also assimilated with a three-dimensional variation (3D-Var) method. Verification, using primarily the estimated rainfall from the weather radar, supplemented by satellite products where needed, shows that radar assimilation significantly enhances convective initiation, storm structure, and peak rainfall placement during the first forecast hours. These results demonstrate that radar-based 4DVAR assimilation can strengthen operational flood early-warning capabilities by providing more reliable rainfall forcing for hydrological and decision-support models. Ongoing work explores integration within multi-sensor workflows, coupling with meteorological forecasting chains, toward operational implementation in Greece.

Key Words: extreme rainfall, WRF, data assimilation, weather radar

How to cite: Katsanos, D., Kalogiros, J., Portalakis, P., Roukounakis, N., and Retalis, A.: Validation of the Precipitation Nowcasting for selected cases in Greece, using weather radar data assimilation, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-16345, https://doi.org/10.5194/egusphere-egu26-16345, 2026.

As climate change intensifies the frequency and magnitude of extreme precipitation, the demand for observation systems capable of accurately capturing short-duration, high-intensity events is increasing, while the limitations of existing frameworks are becoming more apparent. Geostationary (GEO) satellites play a pivotal role in precipitation monitoring due to their high temporal continuity, with the Korean Geo-Kompsat-2A (GK-2A), launched in 2018, providing continuous observations via its Advanced Meteorological Imager (AMI). However, most GEO-based precipitation products rely primarily on infrared (IR) observations, which estimate surface rainfall indirectly from cloud-top radiative properties. Because GEO-based IR precipitation retrievals infer rainfall indirectly from cloud-top signals, a structural limitation arises when cloud-top properties become decoupled from near-surface precipitation processes. This motivates a systematic evaluation of the performance and applicability of GEO IR-based precipitation products under diverse environmental conditions.

In this study, the performance characteristics of the GK-2A precipitation product were evaluated using five years of data (2020–2024) over South Korea, compared against observations from 98 Automated Surface Observing System (ASOS) stations. Quantitative evaluation was conducted for hourly and daily accumulated precipitation using the correlation coefficient (R), Kling–Gupta efficiency (KGE), and unbiased RMSE (ubRMSE), while categorical detection performance was assessed using Accuracy, probability of detection (POD), and false alarm ratio (FAR). Analyses were performed separately for rainy and non-rainy seasons and further stratified by environmental conditions, including air temperature, humidity, cloud fraction, coastal proximity, and terrain ruggedness index (TRI). The microwave-based GPM IMERG product was used as a reference to contextualize the behavior of the IR-based GK-2A estimates.

Results indicate that GK-2A generally exhibits lower correlation and higher error than GPM IMERG, with performance differences becoming more pronounced under specific environmental conditions. Notably, under low temperature and humidity conditions and in coastal regions, GK-2A shows statistically significant performance degradation (p<0.01), characterized by reduced correlation and increased estimation error. In contrast, GPM IMERG maintains relatively stable performance across the same environmental regimes, suggesting that the observed degradation in GK-2A is closely linked to conditions under which cloud-top radiative signals inadequately represent surface precipitation.

By identifying environmental regimes associated with systematic performance degradation, this study clarifies the limitations of GEO IR-based precipitation estimation. The GK-2A case study provides insights applicable to other GEO IR precipitation products and highlights the need for algorithm refinement and multi-sensor integration strategies, particularly incorporating microwave observations, to improve the robustness of high-frequency satellite-based precipitation monitoring under changing climate conditions.

Key Words : Climate change; rainfall intensity; extreme precipitation; ground-based observation; ASOS; GEO-KOMPSAT-2A (GK-2A) satellite

This work was supported by the National Research Foundation of Korea(NRF) grant funded by the Korea government(MSIT) (RS-2025-23523230).

How to cite: Park, S. and Kim, S.: Reliability Assessment and Applicability Analysis of Geostationary Satellite-Based Precipitation Observations for Climate Change Response , EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-16885, https://doi.org/10.5194/egusphere-egu26-16885, 2026.

Accurate near-real-time precipitation estimates are essential for hydrometeorological applications, but are largely limited to regions equipped with ground-based observation networks or rely on infrequent overpasses of low-Earth-orbiting satellites. Geostationary satellites (GEOs) provide continuous, large-scale observations of the atmosphere and surface, offering valuable but indirect information on precipitation. Near-real-time products derived from GEOs face challenges in capturing the occurrence and spatiotemporal variability of rainfall.

We present a conditional Generative Adversarial Network (cGAN) designed to derive quantitative precipitation estimates (QPE) from MSG SEVIRI data. The deep learning model learns the complex, nonlinear relationships between multi-spectral satellite data and surface precipitation. Via the cGAN architecture, with its discriminator, the model is able to predict realistic precipitation fields which also include high and extreme rainfall rates. The model is also able to produce an ensemble of QPE realizations. Model training is done with the high-resolution (1km and 5-minute, aggregated to SEVIRI-resolution) weather radar data RADKLIM-YW in Germany where model performance is also validated. Compared to PDIR-now, from the PERSIANN-family of GEO rainfall products, it shows significant improvement, e.g. PCC increased from 0.32 to 0.47, FARatio decreased from 0.66 to 0.50, POD increased from 0.39 to 0.62. In this contribution we explain the model architecture and show a validation spanning multiple months of data, as well as selected case studies. Furthermore, we discuss planned extensions to additional datasets and the application to the full SEVIRI disc.

How to cite: Janner, S., Glawion, L., Polz, J., and Chwala, C.: Quantitative Precipitation Estimation from SEVIRI IR Data Using Generative AI, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-17336, https://doi.org/10.5194/egusphere-egu26-17336, 2026.

Official rain gauge networks are usually too sparse to capture the spatio-temporal variability of precipitation. To increase network density and thus improve quantitative precipitation estimates, data from crowdsourced personal weather stations (PWS) can be deployed. As these gauges are not professionally placed and maintained, a thorough quality control (QC) prior to the application of PWS data is essential. Although there are currently no standards and guidelines on the QC of rainfall data, two open-source QC frameworks have been developed in recent years. Those are: first, the pypwsqc package (Chwala et al., 2026), which was developed in particular as QC for PWS networks and includes algorithms developed by de Vos et al. (2019) and Bárdossy et al. (2021); and second, RainfallQC, which covers the GSDR-QC framework developed by Lewis et al. (2021). Those QC frameworks are published as Python packages and include several modular methods, filters or checks that can be applied either individually or as a whole framework. 

In this case study, we will explore whether a merged QC approach that combines checks from both frameworks yields better results than the single application of any framework. For this intercomparison, we exploit high-temporal resolution data from a dense network of 12 reliable rain gauges, and around 300 PWS from Reutlingen, Germany. The PWS output of the best QC approach will then be benchmarked against data from nearby professional gauges using precipitation sums and maxima for single events as well as the whole investigation period.

Our results suggest best practices for carrying out QC on rainfall data from PWS, and for different types of rainfall events. We suggest that developing, maintaining and continuously improving open-source QC algorithms supports the use of PWS data in hydrological research.

References

de Vos, L. W., Leijnse, H., Overeem, A., and Uijlenhoet, R.: Quality control for crowdsourced personal weather stations to enable operational rainfall monitoring, Geophysical Research Letters, 46, 8820–8829, 2019. DOI:10.1029/2019GL083731

Bardossy, A., Seidel, J., El Hachem, A.:The use of personal weather station observations to improve precipitation estimation and interpolation, Hydrology and Earth System Sciences, 25, 583-601, 2021. https://doi.org/10.5194/hess-25-583-2021

Chwala, C. et al.: Open-source tools for processing opportunistic rainfall sensor data: An overview of existing tools and the new opensense software packages poligrain, pypwsqc and mergeplg. Submitted to  Hydrology and Earth System Sciences, 2026. 

Lewis, E., Pritchard, D., Villalobos-Herrera, R., Blenkinsop, S., McClean, F., Guerreiro, S., Schneider, U., Becker, A., Finger, P., Meyer-Christoffer, A., Rustemeier, E., Fowler, H. J.: Quality control of a global hourly rainfall dataset, Environmental Modelling & Software, 144, 2021. https://doi.org/10.1016/j.envsoft.2021.105169

How to cite: Zulkarnaen, D., Keel, T., Mohammed, A., Green, A., Chwala, C., and Seidel, J.: Quality Control Algorithms for Precipitation Data - An Intercomparison using Personal Weather Stations, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-17573, https://doi.org/10.5194/egusphere-egu26-17573, 2026.

Flood risk is influenced by a complex interplay between many climatic and non-climatic factors. Among these, heavy precipitation events stand out as one of the primary drivers of flooding. Therefore, the availability and accuracy of precipitation datasets are essential for reliable assessment of flood risk. This study undertakes a comparative analysis of several precipitation products for selected historical large flood events across Canada. The products under investigation include the satellite-based GPM IMERG product, the ERA5-Land reanalysis product, and the Daymet product, which is used as a reference. Since snowfall is frequent and snowmelt is a main driver of flood events in many parts of Canada, our analysis is extended to compare the precipitation products considering surface conditions; i.e. surfaces with and without snow and ice. The evaluation employs a combination of categorical and statistical metrics to assess the accuracy and reliability of the precipitation products. Categorical metrics include the probability of detection, false alarm ratio, and Heidke skill score. Statistical measures such as the correlation coefficient and volume bias are also analysed. These metrics are analysed as functions of precipitation rate, precipitation phase, and surface type. The outcomes of this analysis are anticipated to offer valuable insights for flood modelling studies focused on Canada. Furthermore, the results are expected to provide constructive feedback to algorithm developers, supporting the enhancement of precipitation products, particularly in regions dominated by snow.

How to cite: Yilmaz, K. K., Salinas, J., Bharathi, A., and Otta, K.: Evaluation of Satellite-based and Re-analysis Precipitation Products over Canada, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-21151, https://doi.org/10.5194/egusphere-egu26-21151, 2026.

Weighing precipitation gauges are increasingly being used for ground-based precipitation monitoring because of their greater accuracy compared to tipping bucket gauges, especially in cases of high precipitation intensity and when measuring solid precipitation. Another advantage highlighted by manufacturers of weighing gauges is their lower maintenance requirements. For automatic precipitation measurement GeoSphere Austria currently uses tipping bucket gauges and weighing gauges, the latter with two orifice sizes of 400 cm² and 500 cm². Each of the gauges is equipped with a precipitation detection instrument.

Despite their good performance characteristics, weighing precipitation gauges are sometimes subject to errors, which can be divided into two groups. The first group includes, for example, data outside the measurement range or other errors related to high- or low-amplitude noise, such as incorrect precipitation measurements caused by decanting and/or refilling of the bucket, as well as temperature- or wind-induced errors due to inappropriate noise filtering by the gauge software (spurious precipitation). The other group of errors is related to missing precipitation recordings, mainly caused by internal problems with the gauge-software – this is the opposite case to spurious precipitation measurements, where the internal filter is too restrictive and thus discards the weight gain during a precipitation event.

During quality control process, the data should be corrected for both types of errors wherever possible. The quality control and correction of high-amplitude noise or spurious precipitation values can be performed using appropriate algorithms that are part of the standard check routines (out-of-range tests, internal consistency checks, etc.). Correcting the second type of error (missing precipitation data) is more difficult because the algorithms underlying the data generation by the gauge are unknown (black-box) and there is no way to recover the lost data.

In cases where the internal software filtering is too aggressive, raw bucket-level data could be used to provide an estimate of the missing precipitation amount. In an attempt to account for this source of error at the gauge level, we have developed an automated procedure that combines the change in bucket weight recorded by weighing gauge with measurements from an independent instrument (precipitation monitor) based on 1-min data to filter out mechanical noise and estimate the amount of precipitation that is not recorded by the gauge software. The idea behind this algorithm is that it should be an additional decision-making support for quality control of daily precipitation data in order to obtain precipitation information lost due to software-related issues with the precipitation gauge.

How to cite: Filipovic, N.: Quality control of weighing precipitation gauge measurements, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-21391, https://doi.org/10.5194/egusphere-egu26-21391, 2026.

The launch of the ESA Arctic Weather Satellite Path Finder Mission (AWS-PFM), forerunner of the EUMETSAT EPS-Sterna mission, equipped with a cross-track scanning radiometer (Microwave Radiometer, MWR) which covers frequency between 50 and 325 GHz, represents an important improvement in satellite meteorology. The MWR represents a significant innovation in microwave radiometry, due to its four channels in the 325.15 GHz band offering enhanced sensitivity to cloud ice, thus enabling precise cloud observation.Exploiting coincident overpasses over precipitation events between the AWS and spaceborne radars, such as the Dual-frequency Precipitation Radar (DPR) onboard the NASA/JAXA GPM-CO mission and the Cloud Profiling Radar (CPR) onboard the ESA/JAXA EarthCare mission, can improve our understanding on the relationship between the cloud structure and the signal observed by the radiometer.    

This work presents  a case study concerning a nearly coincident overpass of GPM-CO and AWS-PFM over Hurricane Melissa, a tropical cyclone that developed into a Category 5 during the 2025 Atlantic season. The observations took place on October 30, 2025, when Melissa had weakened to Category 2. The possibility to observe this type of event combining dual-polarization microwave (PMW) channels — available from the GPM Microwave Imager — with the sub-mm channels — available from the AWS-PFM MWR — as well as the measurements and precipitation profiles available from the DPR provides unprecedented potential for improving our understanding of the dynamics and microphysics processes in tropical cyclones. An analysis combining DPR observations and GMI brightness temperature (TB) is carried out based on our previous work regarding the analysis of Mediterranean tropical-like cyclonic events, such as Medicane Ianos, classified as category 1 hurricane at its peak of intensity. In addition, the combination of multi-channel measurements from AWS-PFM MWR reveal the added value of the sub-mm channels at 325.15 GHz to relate the cloud top structure with precipitation features. The comparison between Medicane Ianos and hurricane Melissa shows remarkable similarities at the time of the GPM-CO/AWS-PFM overpass. In both cases, very high Ku-band radar reflectivity values (around 50 dBZ) are associated with very intense precipitation (around 100 mm/h), which does not correspond to extreme TB features usually observed in the presence of strong updrafts sustaining large frozen hydrometeors at the upper levels. This indicates that, even during extremely intense cyclonic phenomena, the development of intense convective cores is limited. 

This analysis is ancillary to the future launch of the EPS Sterna mission, a constellation of AWS-like small satellites, designed to improve weather forecasts by providing global measurements of atmospheric temperature, humidity profiles as well as cloud and precipitation features with frequent revisit times.

How to cite: Camplani, A., Sanò, P., Casella, D., D'Adderio, L. P., Sebastianelli, S., D'Armiento, D., Soncin, L., and Panegrossi, G.: Assessing the Contribution of the Arctic Weather Satellite to Improved Observation of Extreme Cyclonic Events: the hurricane Melissa Case Study, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-20360, https://doi.org/10.5194/egusphere-egu26-20360, 2026.

São Miguel Island (Azores archipelago - Portugal), is located in the North Atlantic and exhibits high spatial and temporal variability in rainfall, strongly controlled by its volcanic morphology and the influence of large-scale atmospheric circulation.The analysis of rainfall trends on São Miguel Island was conducted at annual and seasonal scales using 17 rainfall series (1978/79–2019/20), applying non-parametric statistical methods, namely the Mann–Kendall test to assess trend significance and Sen’s slope estimator to quantify trend magnitude. The analysis reveals a clear predominance of negative trends in both annual and seasonal rainfall, with marked spatial heterogeneity. Statistically significant trends are mainly concentrated in autumn and winter, the seasons accounting for the largest fraction of annual rainfall. Autumn emerges as the season with the highest number and magnitude of negative trends, indicating a consistent transition toward progressively drier conditions. At several rainfall stations, annual trends exceed −20 mm/year, reaching maximum values of −31.6 mm/year at high-altitude sites. These rainfall stations also exhibit significant decreases across multiple seasons, indicating a persistent weakening of the rainfall regime throughout the study period.The relationship between rainfall and the NAO shows a negative annual correlation, with a stronger seasonal signal during autumn. Several stations present statistically significant correlations, indicating that positive NAO phases are associated with reduced rainfall on São Miguel Island. This relationship is particularly consistent in autumn, suggesting that the intensification and persistence of atmospheric patterns associated with positive NAO phases have contributed substantially to the observed negative trends. In contrast, winter correlations are weaker and spatially less coherent, while in spring and summer the influence of the NAO is residual.Overall, the results confirm the dominant role of the NAO as the primary driver of interannual variability and recent rainfall trends on São Miguel Island, highlighting a drying signal in an insular environment that is highly sensitive to changes in North Atlantic atmospheric circulation.

How to cite: Silva, R. F., Marques, R., Zêzere, J. L., and Fragoso, M.: Rainfall Trend Analysis and Its Relationship with the North Atlantic Oscillation on São Miguel Island (Azores, Portugal), EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-10384, https://doi.org/10.5194/egusphere-egu26-10384, 2026.

Convective precipitation is a key component of the hydrological cycle and a major driver of extreme rainfall, flash floods, and other high-impact weather events. Under climate warming, changes in the thermodynamic environment are expected to affect the intensity, spatial structure, duration, and frequency of convective storms. This study investigates long-term changes in convective precipitation over the Czech Republic using time series from 19 observation stations covering the period 1982–2021.

Precipitation totals were classified into convective and stratiform components using an algorithm based on SYNOP reports. The analysis focuses on the warm half of the year (April–September), when convective precipitation dominates. We examine six precipitation characteristics: total precipitation, number of rainy days, rain intensity index, 98th percentile of daily precipitation (P98), seasonal maximum, and the convective fraction. Trends are estimated using Sen’s slope, and their statistical significance is assessed with the Mann–Kendall test. Positive trends are found for all characteristics except the rain intensity index.

In addition, we analyse days with heavy convective precipitation and their relationship to atmospheric circulation. Heavy convective precipitation is defined as a convective precipitation amount exceeding the mean P98 threshold over the study period. Atmospheric circulation types are classified using the Jenkinson and Collins (1977) method. This approach identifies circulation types based on three indices: flow direction, strength, and vorticity. Our results show that heavy convective precipitation most frequently occurs under cyclonic, northwesterly, northerly, and westerly circulation types.

How to cite: Beranova, R. and Rulfova, Z.: Changes in convective precipitation characteristics during the warm season in the Czech Republic, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-10599, https://doi.org/10.5194/egusphere-egu26-10599, 2026.

Raindrop motion can induce measurable high-frequency fluctuations (“scintillations”) in the variance and power spectral density (PSD) of received power from microwave links. This phenomenon was observed in earlier research along with turbulence-induced scintillations. The rain-induced scintillation signature may provide insight into rainfall dynamics at fine spatiotemporal scales (sub-second; meters). Here, we analyze a 26 GHz, 2.2 km microwave-link dataset collected in Wageningen. Using a 20 Hz sampling rate, we compute variance-normalized PSDs over 30 s windows and stratify them by crosswind, using measurements from a weather station located 3 km from the link. We find a roughly monotonic relationship between the integrated spectral power in the 9–10 Hz band and the path-averaged rainfall rate measured by disdrometers. However, substantial unexplained variability remains. Our findings indicate that crosswind effects alone may be insufficient to fully account for the observed variability in the signal. Future work will focus on improved characterization of the local crosswind field and analyzing other rainfall characteristics (e.g. parameters of the raindrop size distribution).

How to cite: Wang, P., Droste, A., Schleiss, M., and Uijlenhoet, R.: Linking Spectral Power to Path-averaged Rainfall Rate in Microwave Link Data, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-14263, https://doi.org/10.5194/egusphere-egu26-14263, 2026.

Quantitative precipitation estimates (QPE) derived from weather radars provide spatially continuous rainfall information but are affected by systematic and random uncertainties, e.g. calibration errors, beam blockage, vertical profile effects, and range-dependent biases. A well-established approach to mitigate these limitations is the adjustment of radar-based precipitation using ground-based reference observations. While rain gauges remain the most common reference, commercial microwave links (CMLs) from cellular communication networks offer a promising complementary source of near-surface rainfall information with high spatial coverage and temporal resolution.

Here, we present the development of the flexible Python-based framework pyRADMAN designed to support operational and research-oriented radar adjustment using multiple types of ground sensors at the Deutscher Wetterdienst. The framework enables preprocessing, configurable selection, and combination of different observations, including both rain gauges and CML attenuation-derived rainfall estimates, with radar data. The system ingests radar data from 17 radar sites, approximately 1500 rain gauges available at DWD, and attenuation data from about 4500 CMLs. A continuous CML data transfer from Ericsson to DWD has been established with a latency of less than 2 minutes, enabling the generation and assessment of near-real-time CML-adjusted radar products. pyRADMAN can be operated in routine mode to provide adjusted QPE products, or in recalculation mode for systematic evaluation and method development.

The applied adjustment approach follows the established principles of the operational RADOLAN adjustment scheme. Additional experiments with radar preprocessing, CML processing strategies and adjustment methods were conducted. We demonstrate the feasibility and performance of radar adjustment that goes beyond the recent operational system by using different sensor configurations including CMLs and a fine temporal resolution. Results are presented for an evaluation period covering July 2023 to December 2024, highlighting the potential benefits and challenges of incorporating CML data for near-real-time radar QPE adjustment in an quasi-operational environment.

How to cite: Wenzel, M., Chwala, C., Maximilian, G., and Winterrath, T.: A Continuously Operating Python Framework for Country-Wide Near-Real-Time Weather Radar Adjustment in Germany Using Rain Gauges and Commercial Microwave Links, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-14108, https://doi.org/10.5194/egusphere-egu26-14108, 2026.

Commercial microwave links (CMLs) are increasingly used as opportunistic sensors for precipitation monitoring, providing high spatiotemporal coverage in urban and regional environments. However, the accuracy of attenuation-based rainfall retrieval from CMLs is strongly affected by systematic fluctuations in the received signal level (RSL) baseline, which often exhibits a pronounced 24-hour periodicity even under dry conditions. The physical origin of this periodic baseline variation remains unclear and represents an important source of uncertainty in precipitation measurements.

In this study, we conduct a controlled outdoor experiment to identify the dominant driver of RSL baseline fluctuations. Targeted thermal perturbations were applied to the outdoor units (ODUs) of operational CMLs, while RSL, receiver-side ODU internal temperature, and ambient air temperature were synchronously recorded. By actively modifying the thermal behavior of the receiver ODUs, we demonstrate that the periodic variation of receiver ODU internal temperature is the primary cause of the RSL baseline fluctuation. When the internal temperature periodicity was disrupted, the corresponding RSL periodicity was significantly weakened, and the apparent correlation between RSL and air temperature disappeared. In contrast, heating applied only to transmitter-side ODUs or insufficient thermal perturbation produced no observable effect.

These findings provide the first experimental evidence that receiver-side instrumental thermal dynamics, rather than atmospheric variability along the propagation path, govern the periodic RSL baseline fluctuation in CML observations. The results identify a key source of bias in CML-based precipitation retrieval and offer a physical basis for improving baseline correction and uncertainty characterization in opportunistic rainfall measurements.

How to cite: Zheng, X., Zhou, J., Ma, D., Qin, Y., and Fu, J.: Experimental identification of receiver-side thermal effects on RSL baseline fluctuations in CML-based precipitation measurements, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-1715, https://doi.org/10.5194/egusphere-egu26-1715, 2026.

The summer months in southeastern Austria are often characterized by severe rainfall from heavy thunderstorms. These events typically unfold rather quickly, with only a few minutes to hours between the formation of the first clouds and the end of the event. Though the intense precipitation during these thunderstorms often results in severe damage, it is still difficult to predict. Deepening our knowledge about the life cycle of such events, from formation to dissipation, is therefore crucial to increasing natural hazard resilience and improving forecasting skills.

Here, we use high-resolution observational data provided by the WegenerNet 3D Open-Air Laboratory for Climate Change Research (WEGN3D Open-Air Lab) located around Feldbach, Austria, to investigate the life cycle of 94 heavy rainfall events. With its 156 ground stations, one X-band radar, two radiometers, and six Global Navigation Satellite System (GNSS) stations, the WEGN3D Open-Air Lab provides high-resolution observations of key atmospheric parameters. In the study, we track 10 atmospheric parameters that are closely linked to heavy precipitation. This gives us insights into characteristic features of the different stages of the precipitation life cycle of small-scale rainfall events.

Starting with the 8 h before the event (i.e., formation stage), we identify distinct features and patterns in air temperature, integrated water vapor, liquid water path, and wind speed that are directly linked to the arrival of the first storm clouds. In the hours of the actual rainfall event (i.e., precipitation stage), the highly localized character of these events is clearly visible in the spatial variability of temperature, liquid water path, and cloud cover. The precipitation triggers a localized cooling effect, which is reflected in a strong correlation between precipitation amount and 2 m air temperature during the event. Subsequently, the integrated water vapor development during the event is also driven by the localized rainfall. In the 16 h after the event (i.e., dissipation stage), we observe the slow return of the atmospheric parameters to pre-event conditions.

The findings of our study are well in-line with the expected physical processes connected to small-scale rainfall extremes. Furthermore, we also demonstrate the WEGN3D Open-Air Lab’s skill to monitor heavy rainfall events and their characteristics in high spatial and temporal resolution. This illustrates the dataset’s high potential for applications in the improvement and verification of weather and climate models.

How to cite: Haas, S., Kvas, A., and Fuchsberger, J.: High-resolution observation-based precipitation life cycle analysis of heavy rainfall events in the southeastern Alpine forelands, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-1818, https://doi.org/10.5194/egusphere-egu26-1818, 2026.

Using high‐resolution hourly precipitation observations over China from 2010 to 2024, this study investigates extreme precipitation from a multi‐time‐scale synergistic perspective, with emphasis on its temporal structure and concentration characteristics. Extreme precipitation series at 1-, 3-, 6-, 12- and 24-hour accumulation scales were constructed to reveal regional differences in persistence and explosiveness. Results show that extreme precipitation in North China is dominated by short‐duration intense rainfall, while in the northern Sichuan Basin it is mainly characterized by long‐lasting events, and South China and the Yangtze–Huaihe region exhibit mixed features. Sub‐daily contribution analysis indicates that, in most parts of central and eastern China, the major portion of 24‐hour extreme precipitation is concentrated within the first three hours, highlighting the dominant role of short‐lived mesoscale convective systems. An Extreme Concentration Index (ECI) is further proposed by integrating actual precipitation contributions with climatic background thresholds, enabling quantitative classification of temporal concentration. The spatial pattern of ECI reveals pronounced geographical differences in the temporal structure of extreme precipitation and shows strong relevance to different disaster risk types. Based on ECI classification, region‐specific conceptual forecasting models and refined prediction strategies are developed, providing an effective scientific framework for improving extreme precipitation forecasting and risk prevention.

How to cite: He, L. and Fu, J.: Multi-Scale Synergistic Characteristics and Temporal Structure of Extreme Precipitation over China, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-3422, https://doi.org/10.5194/egusphere-egu26-3422, 2026.

Precise rainfall estimation is highly essential for investigating water availability, evaluating weather hazards, and understanding rapid climate variations in urban ecosystems. Accurate runoff response is crucial for land use planning, which requires high spatiotemporal precipitation observations, particularly in urban hydrology for groundwater management and the design of efficient drainage systems. Although a rain gauge provides accurate rainfall measurements at a particular location on the surface, it often lacks the spatial extent of rainfall distribution, depending on the gauge network and the complexity of the terrain. Moreover, the rain gauge accumulates the rainwater and records an observation until the minimum threshold of 0.2 mm for rainfall detection is reached, which might miss the precise starting time of the rain event. 

The Weather radar provides a higher spatiotemporal resolution compared to rain gauge monitoring, which tracks precipitation over a larger region at regular spatial and temporal intervals, with an estimate of instantaneous rainfall intensity. X-band weather radar satisfies the need for higher spatiotemporal observation with more accurate rainfall estimates for precise runoff modelling in comparison to S and C-band radars, but at the cost of greater signal attenuation due to its larger operating frequency. X-band radar suffers from the limitation of overshooting for low-lying clouds relative to its sampling volume, which worsens with the increasing range in proportion to the radar elevation angle. X-band radars are also prone to errors resulting from non-meteorological echoes, reflections from ground clutter, and the cone of silence above the maximum elevation angle that causes the rain cells looming above the radar antenna in the zenith direction to remain undetected by the weather radar. Micro rain radar (MRR) is a vertically pointed, specialised, low-cost radar that can continuously measure the drop size distribution and, hence, rainfall rates at different vertical ranges with high resolution. MRR provides fine-scale vertical rainfall characteristics, which can effectively adjust the X-band radar estimates for vertical layers at various elevation angles.

In this research, we have developed a model to perform the bias correction in the rainfall rates of X-band weather radars using the MRR rainfall observation as ground truth. A feed-forward neural network is implemented to improve precipitation estimates from X-band weather radars, utilising rainfall rate, horizontal reflectivity, and specific differential phase as input features. The MRR observations from multiple range gates are averaged over the vertical extent of the X-band radar beam in order to align with the mean rainfall rates from X-band weather radar. The MRR rainfall estimate is pre-processed to mitigate bias by removing outliers caused by evaporation/wind effects, melting particles, or non-meteorological objects, and further verified against collocated rain gauge observations to identify days with actual rainfall events. Thereafter, the rainfall rates at different altitudes from the X-band weather radar nearest to the MRR location are fed to the neural network model as inputs, while the averaged MRR observations from the corresponding range gates are used as the ground truth for training the model. This approach enables bias correction and improves precipitation estimates, particularly across vertical atmospheric layers.

How to cite: Naha Biswas, A. and Hashemi, H.: Augmentation of X-Band Radar Precipitation Estimates with Micro Rain Radar Observations using Machine Learning, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-5164, https://doi.org/10.5194/egusphere-egu26-5164, 2026.

We introduce Version 3 (V3) of the gridded near real-time Multi-Source Weighted-Ensemble Precipitation (MSWEP) product—the first fully global, machine learning-powered precipitation (P) dataset, developed to meet the growing demand for timely and accurate P estimates amid escalating climate challenges. MSWEP V3 provides hourly data at 0.1° resolution from 1979 to the present, continuously updated with a latency of approximately two hours. Development follows a two-stage process. First, baseline P fields are generated using machine learning model stacks that integrate satellite- and (re)analysis-based P and air-temperature products, along with static variables. The models are trained using hourly and daily observations from 15,959 P gauges worldwide. Second, these baseline P fields are corrected using daily and monthly gauge observations from 57,666 and 86,000 stations globally, using a method that accounts for gauge proximity, reporting times, inter-gauge dependencies, and correlation lengths. To assess MSWEP V3's baseline performance, we evaluated 19 (quasi-) global gridded P products—including both uncorrected and gauge-based products—using observations from an independent set of 15,958 gauges excluded from the first training stage. The MSWEP V3 baseline achieved a median daily Kling-Gupta Efficiency (KGE) of 0.69, outperforming all evaluated products. Other uncorrected products achieved median KGE values of 0.61 (ERA5), 0.46 (IMERG-L V7), 0.38 (GSMaP V8), and 0.31 (CHIRP). Notably, the MSWEP V3 baseline also outperformed several gauge-based products, including IMERG-F V7 (0.62), CPC Unified (0.54), and CHIRPS (0.36). Using leave-one-out cross-validation, the daily gauge correction was found to improve the median daily correlation by 0.09, constrained by the already strong baseline performance. We anticipate that MSWEP V3 will substantially advance data-driven decision-making in hydrology and climate science, by enabling more reliable monitoring, forecasting, and management of water-related risks in a variable and changing climate.

How to cite: Beck, H., Wang, X., Alharbi, R., Baez-Villanueva, O., Miralles, D., Ma, J., Xu, S., McCabe, M., Pappenberger, F., van Dijk, A., McVicar, T., Karthikeyan, L., Fowler, H., Pan, M., and Gebrechorkos, S.: MSWEP V3: Machine Learning-Powered Global Precipitation Estimates at 0.1° Hourly Resolution (1979–Present), EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-9714, https://doi.org/10.5194/egusphere-egu26-9714, 2026.