OSA1.4

The crossing-point forecast (CPF) is a new concept in the field of probabilistic forecasting. A CPF is defined by the intersection between a forecast cumulative distribution and the corresponding climatology distribution. Focusing on this intersection point, a probabilistic forecast is summarized into a single number conveying information about a "probabilistic worst-case scenario" with respect to climatology. Is the predicted chance of suffering a loss, due to the occurrence of an (exceedance) event, higher than that event’s climatological frequency? The crossing-point forecast indicates the limit case for which the answer is positive.

The outcome corresponding to a CPF, called “crossing-point observation”, is directly related to the return period of the event that materializes. A simple error function that applies to the forecast and observed crossing-points is formulated. The resulting score is closely related to the diagonal score: it is proper and equitable which makes its application appealing for the comparison of competing forecasts. The proposed scoring function is consistent for the crossing-point forecast in a similar way as the root mean squared error is consistent for the distributional mean forecast or the mean absolute error is consistent for the 50%-quantile forecast.

In weather forecasting, the information provided by CPF could be highly relevant for vulnerable users and more generally for users with interest for high-impact events. We propose here a comparison with the Extreme Forecast Index (EFI) using the ensemble forecast of the Integrated Forecasting System run at ECMWF and the corresponding model climatology. The EFI is designed to provide forecasters with general initial guidance on potential extreme weather events. Both EFI and CPF are derived using the same ingredients which makes their comparison particularly relevant. Based on case studies and verification metrics, we illustrate the complementarity of the two types of forecasts.

How to cite: Ben Bouallegue, Z.: On the crossing-point forecast, EMS Annual Meeting 2022, Bonn, Germany, 5–9 Sep 2022, EMS2022-644, https://doi.org/10.5194/ems2022-644, 2022.

Extreme precipitation events pose a great threat to society, the economy, and the environment with devastating consequences like floods and landslides. Improved forecasting of this hazard, combined with adequate Early Warning Systems, can thus support the mitigation of the associated negative impacts. Here, we propose new forecasting products that substantially improve predictions of extreme precipitation events at different timescales, from short- to medium- and extended-range forecasts. Our study focuses on forecasting extreme rainfall in Calabria, south Italy. This region, located in the central Mediterranean, has a complex and abruptly varying topography that poses additional challenges in precipitation formation and forecasting. For this work we use data from three sources; observational precipitation from the EOBS dataset, modelled atmospheric fields from ERA5 reanalysis, and forecasted atmospheric fields from ECMWF reforecasts for 1 up to 45 days lead time. Based on our analysis we show that different forecasting horizons require different forecasting products. More specifically, bias-correcting the precipitation forecasts for short- to medium-range lead times provides the most informative outputs to end-users. For extended-range forecasts though, this method is not sufficient, as it cannot outperform the reference score based on climatological information. For such extended lead times, it is beneficial to use the large-scale weather variability to infer reliable information about extreme precipitation. This can be accomplished by connecting Mediterranean-wide weather patterns, with the probability of extreme precipitation in Calabria. We present the benefits of the methods, based on long-term statistical analysis using a range of indicators, such as the Brier skill score, and the reliability diagram.

How to cite: Mastrantonas, N., Furnari, L., Magnusson, L., Senatore, A., Mendicino, G., Pappenberger, F., and Matschullat, J.: Forecasting extreme precipitation in the central Mediterranean: Different products for different timescales, EMS Annual Meeting 2022, Bonn, Germany, 5–9 Sep 2022, EMS2022-356, https://doi.org/10.5194/ems2022-356, 2022.

Automated forecasting provides the basis for everyday forecast products used by a wide range of users. Continued progress in numerical weather prediction allows to produce local forecasts with considerable accuracy. To further reduce systematic errors and thereby render such local forecasts more beneficial to users, statistical postprocessing can be employed. While statistical postprocessing can readily be optimized for specific applications, optimization is less straight forward for general-purpose forecasts that are used across a diversity of applications and decisions. This issue is illustrated with ensemble postprocessing for automated precipitation forecasts.

While medium-range precipitation forecasts are often communicated with hourly granularity, beyond the nowcasting range most applications are likely less affected by the precise timing of precipitation. In contrast, (sub-)daily aggregated precipitation  may be a more relevant quantity. In addition, predictability of hourly precipitation is generally very limited days in advance and statistical postprocessing for hourly precipitation forecasts will therefore be strongly affected by the regression-to-the-mean problem (i.e. statistical postprocessing resorts to issuing climatological forecasts in the absence of predictability). To overcome the above issues, we propose a combined postprocessing approach operating on daily aggregated precipitation and hourly fractions of daily precipitation. We present results for a simple disaggregation according to the NWP precipitation and disaggregation according to a separate postprocessing of the hourly fraction of daily totals. The latter approach allows us to correct systematic biases in the diurnal cycle of precipitation occurrence particularly relevant for convective situations in complex topography as is the case in Switzerland. The same approach can be used to extend daily precipitation forecasts into the sub-seasonal to seasonal range.

How to cite: Bhend, J., Spirig, C., Moret, L., and Liniger, M. A.: Multi-resolution postprocessing for precipitation, EMS Annual Meeting 2022, Bonn, Germany, 5–9 Sep 2022, EMS2022-330, https://doi.org/10.5194/ems2022-330, 2022.

Ideally, weather forecasts should be provided for points and not for the large regions represented by global model grid boxes. This requirement can be addressed by post-processing global forecast model output, as in “ecPoint”, an innovative statistical technique developed by ECMWF that uses decision trees and non-local calibration. Products from ecPoint explicitly incorporate the expected sub-grid variability and gridscale bias correction (which both vary according to a diagnosed “grid-box weather type”). Pre-existing 6-h (currently running in CINECA supercomputer facilities in Bologna) and 12-h ecPoint-Rainfall forecasts products are currently being provided by ECMWF in real-time, using shorter range (day 1-15) twice daily predictions at 18km resolution. These probabilistic forecasts have exhibited clear improvements, in both reliability and resolution, relative to the raw model output.

The HIGHLANDER (HIGH-performance computing to support smart LAND sERvices) project started in 2018, funded under the Connecting Europe Facility (CEF) – Telecommunication Sector Programme of the European Union. One of its main goals is the data processing for more intelligent and sustainable management of natural resources and the territory. And one component of this, managed by ECMWF, is exploiting the CINECA HPC capacity to extract maximum benefit from the ecPoint technique. The specific aims here are to improve probabilistic 24-h rainfall and 2m temperature representation in sub-seasonal forecasts and in the ERA5 reanalysis. ecPoint benefits tend to be more significant when working at a lower resolution, so downscaling from 36km in the sub-seasonal forecast and 31 km in ERA5 can in principle deliver greater improvements for users (relative to raw model and reanalysis output) than we have seen in the Medium-Range forecast products.

The final ecPoint sub-seasonal products cover lead times of 16 to 30 days. They provide more reliable climatological representations, for points on the land surface than raw model output (e.g. numbers of dry or wet days, days of freezing temperatures, which are both relevant for agricultural applications). Meanwhile, ERA5 ecPoint products are the first global probabilistic reanalysis products to have incorporated bias-corrected uncertainty information at point scale. ecPoint products for both systems comprise 24-h accumulated rainfall and daily minimum, maximum and mean 2m temperature, for percentiles (1, 2,..99) and (derived from these) probabilities of exceeding certain thresholds. Global outputs are created, but a particular focus is Italy, and surrounding Mediterranean areas, to assist with agricultural planning, to deliver benefits in both economic and resource availability terms.

In this presentation, we will describe ECMWF activities in HIGHLANDER and the methods applied to create the final probabilistic forecast and reanalysis products.

How to cite: Gascón, E., Vintzileos, A., and Hewson, T.: New probabilistic point forecast products ("ecPoint") for sub-seasonal forecasts and the ERA5 reanalysis -  the HIGHLANDER project, EMS Annual Meeting 2022, Bonn, Germany, 5–9 Sep 2022, EMS2022-457, https://doi.org/10.5194/ems2022-457, 2022.