HOPSS-X (HOurly Precipitation Spatial Structures and Xtremes) from MERIDA HRES Reanalysis

HOPSS-X (HOurly Precipitation Spatial Structures and Xtremes) from MERIDA HRES Reanalysis

This dataset provides a database of hourly precipitation spatial structures over Italy, derived from the MERIDA HRES convection-permitting reanalysis (1986–2022), described in Viterbo et al. (2024) https://doi.org/10.1002/met.70011.

The dataset focuses on hourly precipitation structures, identified from each hourly field using a percentile-based clustering technique that accounts for both seasonal and regional variability in precipitation regimes. The clustering threshold is defined as the 50th percentile of all hourly precipitation values exceeding 1 mm/h, calculated separately for each season and grid point. This adaptive method reduces the influence of light, spatially noisy precipitation and ensures selecting spatially continuous precipitation structures that are assumed to have a detectable, and at times significant, impact on human activities, as they do not fall into the category of very light precipitation, given the season and the region under study.

For each hourly field, contiguous grid cells exceeding the threshold were grouped into a single structure, provided that the cluster covered at least five grid cells, to exclude small-scale noise. The detection was performed independently for each hour, without tracking structures over time, in line with the dataset’s focus on the spatial characteristics of hourly precipitation.

For each event, the dataset provides a set of features, including:

• Date and hour of occurrence

• Maximum precipitation intensity (mm/h) and its location

• Intensity-weighted centroid coordinates

• Total event area (number of grid cells)

• Total precipitation volume (mm/h)

• Geometrical shape descriptors: eccentricity, orientation angle, and major axis length, calculated using the minimum enclosing ellipse method.

Over the 37-year analysis period, the dataset contains approximately 160,000 precipitation structures per year. The interannual variability in the annual number of detected structures is around 10%, as quantified by the relative standard deviation across years.

To support diverse user needs, the dataset is provided as individual yearly .txt files (1986–2022); each row represents a different structure, each column a different feature. Additionally, a separate aggregated file is included, containing only the Hourly Precipitation Extremes (HPEs) for the whole period. The HPEs filtering is based on a threshold derived from the local annual maxima of hourly precipitation (RX1hour).

This dataset aims to facilitate research on precipitation extremes, hydrological impacts, and climate change signals over Italy. Further methodological details, validation steps, and scientific analyses—especially regarding the trends and spatial patterns of precipitation extremes over Italy—will be available in the associated scientific article currently under review.