Ability of Austria’s high spatiotemporal resolution of INCA for rainfall erosivity assessment in the main agricultural production zones

Rainfall erosivity is a critical parameter for soil degradation assessment. It is especially important in areas prone to soil erosion due to frequent or intense rainfall events. Rainfall erosivity is typically quantified through the EI30 index, which combines the rainfall's kinetic energy (E) with its maximum 30-minute intensity (I30). Rainfall data with high temporal resolution (e.g., 5-minute) are often limited in availability, particularly across large spatial scales. Gridded rainfall products usually provide half-hourly or hourly data and are frequently used despite the potential biases they introduce into erosivity calculations. In Austria, the introduction of INCA (Integrated Nowcasting through Comprehensive Analysis) has provided an opportunity to enhance rainfall erosivity studies. INCA offers rainfall data at a 15-minute temporal and 1-kilometer spatial resolution. In our study, INCA’s accuracy and limitations in capturing erosive rainfall events, especially extreme events, are carefully evaluated against high-resolution in-situ rainfall station data. Understanding the degree of underestimation or overestimation in EI30 calculations is crucial for applying INCA data in erosion simulations and according to soil conservation strategies. This study focuses on the Main Agricultural Production Zones (MAPZ) in Austria. The study evaluates the performance of INCA in calculating erosivity compared to a dense network of 5-minute resolution rainfall stations. It also investigates the occurrence probability of extreme erosive events using a probabilistic approach, providing insights into long-term erosion risks and regional differences in erosivity patterns. Eventually, the study examines the impact of temporal resolution on erosivity estimates, assessing biases introduced by coarser resolutions (15- and 30-minute) for both mean and extreme rainfall events. Results indicate that INCA overestimates 8.1% of the total event number but underestimates 3.1% of the total rainfall erosivity, with the largest under- and overestimation in east of Austria. On the other hand, the mean annual maximum EI30 was underestimated by 13.6%, and the south showed the most considerable underestimation. It was found that INCA can detect highly erosive events occurring in the in-situ datasets. The underestimation of EI30 sources from the temporally smoothened peak I30 rather than the E of the rainfall events. Long-term extreme EI30 were analyzed using the Generalized Extreme Value (GEV) distribution, revealing that EI30 values increase with longer return periods (e.g., 50 years) and that the southern region exhibits the largest EI30 values, indicating a greater risk of extreme erosive events. On the other hand,  INCA may emphasize more recent, potentially intense rainfall trends, leading to larger return levels. The impact assessment by coarser temporal resolutions on EI30 confirms that the underestimation substantially increases with lower temporal resolution, primarily due to I30 rather than E. Eventually, a 15-minute temporal resolution dataset may lead to acceptable underestimations across our investigated MAPZ; underestimations ranged from 2.8–8.5% in event numbers and 1.0–10.0% in total rainfall erosivity at 15-minute resolution and from 10.2–33.7% and 5.2–26.6%, respectively, at a 30-minute resolution. The results of this study highlight the potential value of INCA data as a practical source for rainfall erosivity assessments, particularly in regions with limited high-resolution rainfall measurements.