UP3.3

Climate modifiers over Croatia such as the Adriatic sea, highly developed orography of Gorski kotar and the Dinarides, and the Slavonian plain cause the great climatic differences between Croatian regions. Related to this, regional weather conditions can differ a lot within the same type of large-scale atmospheric circulation.
Until now, the weather types that are commonly in use in Croatia have been determined according to the subjective Poje’s classification which consists of 29 weather types determined after the distribution of surface pressure and defined for relatively small and diverse areas. Hence, the Croatian area is divided into five regions by expert’s experience: Eastern and Central Croatia, North, Middle and South Adriatic.
The motivation of the here presented study is to improve previous knowledge and practice that is mainly based on the application of Poje’s classification by implementing the results of the COST733 action “Harmonization and Applications of Weather Types Classifications for European Regions”. For the first time, the efficacy of different objective classifications provided by COST733 will be examined utilizing Croatian meteorological station data. This study intends to determine the most suitable combination of main classification parameters such as domain size, number of types, input variables and classification methods. Based on these results, in a subsequent step, an optimized classification that is intended to be the most appropriate to capture precipitation conditions in Croatian regions should be developed.
For evaluating the discriminative power (synoptic skill) of the classifications and the relevance of specific settings several statistical metrics will be used, some of them consider precipitation intensity, while others consider precipitation occurrence/ absence. The assessment of evaluation results takes into account the effect of varying numbers of types and specifies spatial (among regions) and as well temporal (among seasons) variations.
Some preliminary results have shown better classification performance along the Adriatic coast and in the mountainous region than in the more continental parts, as well as for optimization and threshold based methods among other methods. Furthermore, better classification performances are found in the cold part of the year (winter, autumn) and spring compared to summer, as well as for single day classifications than for four days ones. However, there is no clear improvement in the inclusion of additional variables. Statistic metrics calculated from occurrence/ absence of precipitation exhibit larger values (better performance) than those calculated from precipitation intensity. Moreover, results have pointed out the importance of temporal compliance of the datasets.

How to cite: Marinovic, I. and Beck, C.: Application of COST733 Objective Classifications in Croatia, EMS Annual Meeting 2022, Bonn, Germany, 5–9 Sep 2022, EMS2022-302, https://doi.org/10.5194/ems2022-302, 2022.

We present results from atmospheric circulation type classification of climate simulations over Europe using an ensemble of 57 regional simulations from Euro-CORDEX, their 11 global boundary models from CMIP5 as well as the ERA5 reanalysis. We compared frequencies of the different circulation types in the simulations with ERA5 and found that the regional models add value in the summer season. We applied three different classification methods (the subjective Grosswettertypes and the two optimisation algorithms SANDRA and distributed k-means clustering) from the cost733class software and found that the results are not particularly sensitive to choice of circulation classification method.

There are large differences between models. Simulations based on MIROC-MIROC5 and CNRM-CERFACS-CNRM-CM5 show an over-representation of easterly flow and an under-representation of westerly. Comparing the same day in the global and regional models, the downscaled results retain the large-scale circulation from the global model most days, but especially the regional model IPSL-WRF381P changes the circulation more often, which increases the error relative to ERA5. Simulations based on ICHEC-EC-EARTH and MPI-M-MPI-ESM-LR show consistently smaller errors relative to ERA5 in all seasons. The ensemble spread is largest in summer and smallest in winter.

Under the future RCP8.5 scenario, the circulation changes in the summer season, with more than half of the ensemble showing a decrease in frequency of the Central-Eastern European high, the Scandinavian low as well as south-southeasterly flow. There is in general a strong agreement in the sign of the change between the regional simulations and the data from the corresponding global model.

We hope that evaluations like these can provide useful insight for example when selecting regional model simulations for use in climate services.

How to cite: Røste, J. and Landgren, O.: How do large-scale circulation type statistics change after downscaling in the Euro-CORDEX ensemble?, EMS Annual Meeting 2022, Bonn, Germany, 5–9 Sep 2022, EMS2022-295, https://doi.org/10.5194/ems2022-295, 2022.

How to cite: Lehner, S. and Schellander-Gorgas, T.: Implication of ensemble selection and weighting on the performance of regional climate simulations in Austria, EMS Annual Meeting 2022, Bonn, Germany, 5–9 Sep 2022, EMS2022-364, https://doi.org/10.5194/ems2022-364, 2022.

The modes of low-frequency variability of atmospheric circulation (teleconnections) have been studied mostly in winter. Of the four modes dominating over the North Atlantic – European sector (North Atlantic Oscillation (NAO), and East Atlantic (EA) and two Eurasian (EU1, EU2) patterns), only the NAO has been investigated in other seasons. The objective of this study is to describe the development of the North Atlantic – Europen circulation modes throughout the year, that is, to characterize their annual cycle.

The modes are identified in 500 hPa heights in the Northern Extratropics (north of 20°N inclusive) for period 1948-2016. The data source is the NCEP/NCAR reanalysis. The analysis tool is rotated principal component analysis (using covariance matrix and varimax rotation). The key ingredience of our methodology is that the analysis is based on sliding seasons 90 days long, which are moved with a step of 5 days. That is, the entire year is covered by 73 (=365:5) sliding seasons. The number of rotated components (that is, of detected modes) varies from 9 in winter to 13 in summer.

The four standard modes listed above are detected unambiguously from December to March. They change their position and spatial structure in spring, when also other modes appear. Specifically, the centres of the NAO shift westward in spring and northward in summer. The EA pattern loses its zonal character in the warm half year; its southern centre shifts towards Europe and weakens. The behaviour of the Eurasian modes is more complex. Since they describe a moving wave, being in quadrature one with another, their geographical position is to some extent random and their attribution in successive sliding seasons is hardly possible in spring and autumn. The shape of the modes stabilizes in May and remains fairly stable until October when the circulation returns to the winter character. The results document the changes in circulation action centres and their connectivity and show the obvious asymmetry of annual course of modes. One point correlation maps are employed to support the physical realism of the modes detected by principal component analysis.

How to cite: Huth, R. and Pokorná, L.: Annual cycle of modes of low-frequency circulation variability, EMS Annual Meeting 2022, Bonn, Germany, 5–9 Sep 2022, EMS2022-253, https://doi.org/10.5194/ems2022-253, 2022.

Principal Component Analysis (PCA) is a well-known and widely used technique in climatology to identify modes of low-frequency variability of atmospheric circulation. PCA is also used to examine changes in mode patterns over time. We present temporal variability of modes using moving PCA of winter (DJF) monthly mean 500 hPa height anomalies for 20 to 50-year long moving periods with one year step. We compare modes identified for the ensemble mean and ensemble members of the 20CRv2, CERA20, ERA5 and NCEP/NCAR reanalyses. It shows that PCA is sensitive to the period analysed and results can vary considerably if the period is shifted by even one year. Changes in the pattern of a mode are usually interpreted as real changes in atmospheric circulation. However, these changes can be influenced by many aspects, e.g., data quality, length of the analysed period, or settings of PCA. The sudden changes from period to period are more common whit shorter moving periods, while longer periods give more stable results. Some of these sudden changes appear to be more or less random, while others occur in all ensemble members or even in all reanalyses used in the same period. This suggests that comparisons of modes from just a few periods, or even just two periods, may be affected by the presence of a sudden change. Comparing modes from periods that are shifted by only one year, for example, may then provide different results and hence a different interpretation of the shift in the centres of variability.

How to cite: Piskala, V. and Huth, R.: Sensitivity of moving PCA in detecting spatial changes in teleconnections, EMS Annual Meeting 2022, Bonn, Germany, 5–9 Sep 2022, EMS2022-361, https://doi.org/10.5194/ems2022-361, 2022.

Variability of weather conditions in the temperate zone is mainly related to the atmospheric circulation, including the presence of atmospheric fronts. Especially extreme weather phenomena such as wind gusts, heavy precipitation, hail, thunderstorms, and many others are usually occurred with passing of weather front. Climatology of weather fronts is difficult, since they are usually drawn manually by forecasters on maps, therefore the database of them is limited and the process itself is very subjective in its nature. In this study front analysis climatology over Central Europe (5°–30° W and 45°–60° N) is presented. It is based on an objective method.

The key issue in the study was to develop a method for the detection of atmospheric fronts. After many attempts and analyzes, the method based on based on random forest machine learning technique was used to distinguish the fronts. On the basis of a 3-year testing sample (2019-2021) of digitized synoptic maps developed by DWD, a system was built using the AI method.

The digitized weather fronts were used for further analysis based on data from an ERA5 meteorological reanalysis. We used surface data as well as data from 4 pressure levels (925, 850, 750, and 500 hPa) and 16 available meteorological fields that can indicate changes in weather conditions in the selected area. Due to relatively high dependency between original frontal information and automatically detected fronts the next step was performed, i.e., the preparation of fronts database for the entire 30-years 1991-2020. Finally, it enabled to create fronts climatology for the area of interest.

Therefore, the frequency of the occurrence of fronts over Central Europe was assessed. Due to methodological conditions, the analysis covered the number of days with the front over the investigated area. The study was made for an individual sub-areas with the size of 5x5° for the period 1991-2020. The analysis showed a slightly higher frequency of the number of days with a front in the year and in individual seasons in the NW part of the study area. The analysis of the variability of the number of days with fronts in the analyzed multiannual period did not show any significant differences between individual years. Larger differentiation was found only when analyzing the number of days with fronts in individual months.

How to cite: Wypych, A., Ustrnul, Z., Bochenek, B., and Kubacka, D.: Atmospheric fronts climatology over Central Europe, EMS Annual Meeting 2022, Bonn, Germany, 5–9 Sep 2022, EMS2022-563, https://doi.org/10.5194/ems2022-563, 2022.

"ecPoint" is a statistical post-processing technique that anticipates sub-grid variability and estimates biases in numerical weather prediction (NWP) model outputs, de facto downscaling such outputs from grid-box to point-scale. Forecasts are post-processed based on the distribution of errors in short-range forecasts versus rain gauge observations in the calibration period. This talk will focus on the branch of the ecPoint family products that post-processes ECMWF ensemble (ENS) rainfall forecasts. ecPoint applies the "remote calibration" approach, whose concept is based on the fact that the physics of rainfall generation is the same around the globe. The "remote calibration" approach then allows the post-processing of forecasts in location X using past observations in different regions, countries, or continents as far as the rainfall-generating weather type (WT) between location X and the remote locations is similar. "Weather types" are identified at grid-box scale and can be defined with different degrees of precision. For example, one can have a single-WT scenario where every grid-box is post-processed with the same error distribution. Or one can have a multiple-WT scenario where n significantly different error distributions are defined using predictors calculated from NWP variables (e.g. the fraction of convective rainfall in the total precipitation forecasts, the speed of steering winds, CAPE, etc.) or location-related variables (e.g. solar radiation, orography, etc.). Each grid-box forecast is then post-processed with the corresponding error distribution. In spite of verification having shown that multi-WT ecPoint-Rainfall (214 WTs derived from five predictors) provides more reliable and skilful rainfall forecasts than raw ENS (for point-verification, especially in case of extremes, e.g. rainfall greater than 50 mm/12h), such a product requires the investment of resources to manually re-calibrate ecPoint-Rainfall for new model cycles. Although a multiple-WT ecPoint-Rainfall delivers better forecasts, a single-WT product would be cheaper to re-calibrate. So, is it worth maintaining a multiple-WT ecPoint-Rainfall? This talk aims to answer this question. 

How to cite: Pillosu, F. and Hewson, T.: To what extent does the diagnosis of multiple grid-box weather types add value in post-processing ensemble rainfall forecasts?, EMS Annual Meeting 2022, Bonn, Germany, 5–9 Sep 2022, EMS2022-671, https://doi.org/10.5194/ems2022-671, 2022.