TY - JOUR A1 - Ibebuchi, Chibuike Chiedozie A1 - Schönbein, Daniel A1 - Paeth, Heiko T1 - On the added value of statistical post-processing of regional climate models to identify homogeneous patterns of summer rainfall anomalies in Germany JF - Climate Dynamics N2 - A fuzzy classification scheme that results in physically interpretable meteorological patterns associated with rainfall generation is applied to classify homogeneous regions of boreal summer rainfall anomalies in Germany. Four leading homogeneous regions are classified, representing the western, southeastern, eastern, and northern/northwestern parts of Germany with some overlap in the central parts of Germany. Variations of the sea level pressure gradient across Europe, e.g., between the continental and maritime regions, is the major phenomenon that triggers the time development of the rainfall regions by modulating wind patterns and moisture advection. Two regional climate models (REMO and CCLM4) were used to investigate the capability of climate models to reproduce the observed summer rainfall regions. Both regional climate models (RCMs) were once driven by the ERA-Interim reanalysis and once by the MPI-ESM general circulation model (GCM). Overall, the RCMs exhibit good performance in terms of the regionalization of summer rainfall in Germany; though the goodness-of-match with the rainfall regions/patterns from observational data is low in some cases and the REMO model driven by MPI-ESM fails to reproduce the western homogeneous rainfall region. Under future climate change, virtually the same leading modes of summer rainfall occur, suggesting that the basic synoptic processes associated with the regional patterns remain the same over Germany. We have also assessed the added value of bias-correcting the MPI-ESM driven RCMs using a simple linear scaling approach. The bias correction does not significantly alter the identification of homogeneous rainfall regions and, hence, does not improve their goodness-of-match compared to the observed patterns, except for the one case where the original RCM output completely fails to reproduce the observed pattern. While the linear scaling method improves the basic statistics of precipitation, it does not improve the simulated meteorological patterns represented by the precipitation regimes. KW - summer precipitation regions KW - Germany KW - climate models KW - fuzzy classification KW - bias correction Y1 - 2022 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:bvb:20-opus-324122 SN - 0930-7575 VL - 59 IS - 9-10 ER - TY - JOUR A1 - Ibebuchi, Chibuike Chiedozie A1 - Schönbein, Daniel A1 - Adakudlu, Muralidhar A1 - Xoplaki, Elena A1 - Paeth, Heiko T1 - Comparison of three techniques to adjust daily precipitation biases from regional climate models over Germany JF - Water N2 - This study compares the performance of three bias correction (BC) techniques in adjusting simulated precipitation estimates over Germany. The BC techniques are the multivariate quantile delta mapping (MQDM) where the grids are used as variables to incorporate the spatial dependency structure of precipitation in the bias correction; empirical quantile mapping (EQM) and, the linear scaling (LS) approach. Several metrics that include first to fourth moments and extremes characterized by the frequency of heavy wet days and return periods during boreal summer were applied to score the performance of the BC techniques. Our results indicate a strong dependency of the relative performances of the BC techniques on the choice of the regional climate model (RCM), the region, the season, and the metrics of interest. Hence, each BC technique has relative strengths and weaknesses. The LS approach performs well in adjusting the first moment but tends to fall short for higher moments and extreme precipitation during boreal summer. Depending on the season, the region and the RCM considered, there is a trade-off between the relative performances of the EQM and the MQDM in adjusting the simulated precipitation biases. However, the MQDM performs well across all considered metrics. Overall, the MQDM outperforms the EQM in improving the higher moments and in capturing the observed return level of extreme summer precipitation, averaged over Germany. KW - bias correction KW - multivariate quantile delta mapping KW - empirical quantile mapping KW - linear scaling KW - precipitation KW - Germany Y1 - 2022 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:bvb:20-opus-262064 SN - 2073-4441 VL - 14 IS - 4 ER - TY - THES A1 - Ibebuchi, Chibuike Chiedozie T1 - Bias correction of climate model output for Germany T1 - Bias-Korrektur des Klimamodell-Outputs für Deutschland N2 - Regional climate models (RCMs) are tools used to project future climate change at a regional scale. Despite their high horizontal resolution, RCMs are characterized by systematic biases relative to observations, which can result in unrealistic interpretations of future climate change signals. On the other hand, bias correction (BC) is a popular statistical post-processing technique applied to improve the usability of output from climate models. Like every other statistical technique, BC has its strengths and weaknesses. Hence, within the regional context of Germany, and for temperature and precipitation, this study is dedicated to the assessment of the impact of different BC techniques on the RCM output. The focuses are on the impact of BC on the RCM’s statistical characterization, and physical consistency defined as the spatiotemporal consistency between the bias-corrected variable and the simulated physical mechanisms governing the variable, as well as the correlations between the bias-corrected variable and other (simulated) climate variables. Five BC techniques were applied in adjusting the systematic biases in temperature and precipitation RCM outputs. The BC techniques are linear scaling, empirical quantile mapping, univariate quantile delta mapping, multivariate quantile delta mapping that considers inter-site dependencies, and multivariate quantile delta mapping that considers inter-variable dependencies (MBCn). The results show that each BC technique adds value in reducing the biases in the statistics of the RCM output, though the added value depends on several factors such as the temporal resolution of the data, choice of RCM, climate variable, region, and the metric used in evaluating the BC technique. Further, the raw RCMs reproduced portions of the observed modes of atmospheric circulation in Western Europe, and the observed temperature, and precipitation meteorological patterns in Germany. After the BC, generally, the spatiotemporal configurations of the simulated meteorological patterns as well as the governing large-scale mechanisms were reproduced. However, at a more localized spatial scale for the individual meteorological patterns, the BC changed the simulated co-variability of some grids, especially for precipitation. Concerning the co-variability among the variables, a physically interpretable positive correlation was found between temperature and precipitation during boreal winter in both models and observations. For most grid boxes in the study domain and on average, the BC techniques that do not adjust inter-variable dependency did not notably change the simulated correlations between the climate variables. However, depending on the grid box, the (univariate) BC techniques tend to degrade the simulated temporal correlations between temperature and precipitation. Further, MBCn which adjusts biases in inter-variable dependency has the skill to improve the correlations between the simulated variables towards observations. N2 - Regionale Klimamodelle (RCMs) sind Werkzeuge, die verwendet werden, um den zukünftigen Klimawandel auf regionaler Ebene zu prognostizieren. Trotz ihrer hohen horizontalen Auflösung sind RCMs je nach Beobachtung durch systematische Verzerrungen gekennzeichnet, was zu unrealistischen Interpretationen zukünftiger Signale des Klimawandels führen kann. Andererseits ist die Bias-Korrektur (BC) eine beliebte statistische Nachbearbeitungstechnik, die angewendet wird, um die Nutzbarkeit der Ergebnisse von Klimamodellen zu verbessern. Wie jede andere statistische Technik hat BC seine Stärken und Schwächen. Daher widmet sich diese Studie im regionalen Kontext Deutschlands und für Temperatur und Niederschlag der Bewertung der Auswirkungen verschiedener BC-Techniken auf den das RCM-ErtragErgebnis. Die Schwerpunkte liegen auf der Auswirkung von BC auf die statistische Charakterisierung des RCM und auf der physikalischen Konsistenz. Letztere ist, definiert als die räumlich-zeitliche Konsistenz zwischen der systematisch korrigierten Variablen und den simulierten physikalischen Mechanismen, die diese Variable steuern, sowie auf den Korrelationen zwischen der systematisch korrigierten Variablen und anderen (simulierten) Klimavariablen. Fünf BC-Techniken wurden angewendet, um die systematischen Abweichungen in den Temperatur- und Niederschlags-RCM-Ausgaben Ergebnissen anzupassen. Die BC-Techniken sind lineare Skalierung, empirisches Quantil-Mapping, univariates Quantil-Delta-Mapping, sowie multivariates Quantil-Delta-Mapping, das Abhängigkeiten zwischen Standorten berücksichtigt, und multivariates Quantil-Delta-Mapping, das intervariable Abhängigkeiten (MBCn) berücksichtigt. Die Ergebnisse zeigen, dass jede BC-Technik einen Mehrwert bei der Reduzierung der Verzerrungen in den Statistiken der RCM-Ausgabe bringt, und dies, obwohl der Mehrwert von mehreren Faktoren abhängt, wie der zeitlichen Auflösung der Daten, der Wahl der RCM, der Klimavariable, der Region und desr verwendeten Massstabsetrik zur Bewertung der BC-Technik verwendet. Darüber hinaus reproduzierten die rohen RCMs Teile der beobachteten Modi der atmosphärischen Zirkulation in Westeuropa und die beobachteten meteorologischen Temperatur- und Niederschlagsmuster in Deutschland. Nach der BC wurden im Allgemeinen die raumzeitlichen Konfigurationen der simulierten meteorologischen Muster sowie die maßgeblichen großräumigen Mechanismen reproduziert. Auf einer stärker lokalisierten räumlichen Skala änderte der BC jedoch für die einzelnen meteorologischen Muster die simulierte Kovariabilität einiger Gitter, insbesondere für Niederschlag. Bezüglich der Kovariabilität zwischen den Variablen wurde sowohl in Modellen als auch in Beobachtungen eine physikalisch interpretierbare positive Korrelation zwischen Temperatur und Niederschlag im borealen Winter gefunden. Für die meisten Gitterboxen Gitterfelder im Untersuchungsbereich und auch im Durchschnitt änderten die BC-Techniken, die die Abhängigkeit zwischen den Variablen nicht anpassen, die simulierten Korrelationen zwischen den Klimavariablen nicht merklich. Allerdings neigen die (univariaten) BC-Techniken je nach Gitterbox Gitterfeld dazu, die simulierten zeitlichen Korrelationen zwischen Temperatur und Niederschlag zu verschlechtern. Darüber hinaus hat MBCn, das Verzerrungen in der Abhängigkeit zwischen Variablen anpasst, die Fähigkeit, die Korrelationen zwischen den simulierten Variablen gegenüber den Beobachtungen zu verbessern. KW - Bias correction KW - regional climate models KW - Germany KW - physical consistency KW - meteorological patterns KW - Bias-Korrektur KW - Regionale Klimamodelle KW - Deutschland KW - Physikalische Konsistenz KW - Meteorologische Muster Y1 - 2023 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:bvb:20-opus-312647 ER - TY - JOUR A1 - Ibebuchi, Chibuike Chiedozie T1 - Patterns of atmospheric circulation in Western Europe linked to heavy rainfall in Germany: preliminary analysis into the 2021 heavy rainfall episode JF - Theoretical and Applied Climatology N2 - The July 2021 heavy rainfall episode in parts of Western Europe caused devastating floods, specifically in Germany. This study examines circulation types (CTs) linked to extreme precipitation in Germany. It was investigated if the classified CTs can highlight the anomaly in synoptic patterns that contributed to the unusual July 2021 heavy rainfall in Germany. The North Atlantic Oscillation was found to be the major climatic mode related to the seasonal and inter-annual variations of most of the classified CTs. On average, wet (dry) conditions in large parts of Germany can be linked to westerly (northerly) moisture fluxes. During spring and summer seasons, the mid-latitude cyclone when located over the North Sea disrupts onshore moisture transport from the North Atlantic Ocean by westerlies driven by the North Atlantic subtropical anticyclone. The CT found to have the highest probability of being associated with above-average rainfall in large part of Germany features (i) enhancement and northward track of the cyclonic system over the Mediterranean; (ii) northward track of the North Atlantic anticyclone, further displacing poleward, the mid-latitude cyclone over the North Sea, enabling band of westerly moisture fluxes to penetrate Germany; (iii) cyclonic system over the Baltic Sea coupled with northeast fluxes of moisture to Germany; (iv) and unstable atmospheric conditions over Germany. In 2021, a spike was detected in the amplitude and frequency of occurrence of the aforementioned wet CT suggesting that in addition to the nearly stationary cut-off low over central Europe, during the July flood episode, anomalies in the CT contributed to the heavy rainfall event. KW - circulation type (CT) KW - atmospheric circulation KW - Western Europe KW - Germany KW - flood KW - heavy rainfall Y1 - 2022 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:bvb:20-opus-324100 SN - 0177-798X VL - 148 IS - 1-2 ER -