TY  - JOUR
A1  - Maraun, Douglas
A1  - Rust, Henning W.
A1  - Osborn, Tim J.
T1  - The annual cycle of heavy precipitation across the United Kingdom : a model based on extreme value statistics
N2  - The annual cycle of extreme I-day precipitation events across the UK is investigated by developing a statistical model and fitting it to data from 689 rain gauges A generalized extrerne-value distribution (GEV) is fit to the time series of monthly maxima, across all months of the year simultaneously, by approximating, the annual cycles of the location and scale parameters by harmonic functions, while keeping the shape parameter constant throughout the year We average the shape parameter of neighbouring rain gauges to decrease uncertainties. and also Interpolate values of all model parameters to give complete coverage of (lie UK. The model reveals distinct spatial patterns the estimated parameters The annual mean of the location and scale parameter is highly correlated with orography. The annual cycle of the location parameter is strong in the northwest UK (peaking in late autumn or winter) and in East Anglia (where it peaks HI late summer), and low in the Midlands The annual cycle of the scale parameter exhibits a similar pattern with strongest amplitudes in East Anglia The spatial patterns of the annual cycle phase suggest that they are linked to the dominance of frontal precipitation for generating extreme precipitation in the west and convective precipitation in the southeast of the UK The shape parameter shows a gradient from Positive Values in the east to negative values in some areas of the west We also estimate 10-year and 100-year return levels at each rain gauge, and interpolated across the UK.
Y1  - 2009
UR  - http://www3.interscience.wiley.com/cgi-bin/jhome/4735
U6  - https://doi.org/10.1002/Joc.1811
SN  - 0899-8418
ER  - 
TY  - JOUR
A1  - Maraun, Douglas
A1  - Rust, H. W.
A1  - Timmer, Jens
T1  - Tempting long-memory : on the interpretation of DFA results
N2  - We study the inference of long-range correlations by means of Detrended Fluctuation Analysis (DFA) and argue that power-law scaling of the fluctuation function and thus long-memory may not be assumed a priori but have to be established. This requires the investigation of the local slopes. We account for the variability characteristic for stochastic processes by calculating empirical confidence regions. Comparing a long-memory with a short-memory model shows that the inference of long-range correlations from a finite amount of data by means of DFA is not specific. We remark that scaling cannot be concluded from a straight line fit to the fluctuation function in a log-log representation. Furthermore, we show that a local slope larger than alpha=0.5 for large scales does not necessarily imply long memory. We also demonstrate, that it is not valid to conclude from a finite scaling region of the fluctuation function to an equivalent scaling region of the autocoffelation function. Finally, we review DFA results for the Prague temperature data set and show that long-range correlations cannot not be concluded unambiguously
Y1  - 2004
SN  - 1023-5809
ER  - 
TY  - JOUR
A1  - Maraun, Douglas
A1  - Kurths, Jürgen
T1  - Epochs of phase coherence between El Nino/Southern Oscillation and Indian monsoon
N2  - We present a modern method used in nonlinear time series analysis to investigate the relation of two oscillating systems with respect to their phases, independently of their amplitudes. We study the difference of the phase dynamics between El Nino/Southern Oscillation (ENSO) and the Indian Monsoon on inter-annual time scales. We identify distinct epochs, especially two intervals of phase coherence, 1886 - 1908 and 1964 - 1980, corroborating earlier findings from a new point of view. A significance test shows that the coherence is very unlikely to be the result of stochastic fluctuations. We also detect so far unknown periods of coupling which are invisible to linear methods. These findings suggest that the decreasing correlation during the last decades might be a typical epoch of the ENSO/ Monsoon system having occurred repeatedly. The high time resolution of the method enables us to present an interpretation of how volcanic radiative forcing could cause the coupling
Y1  - 2005
SN  - 0094-8276
ER  - 
TY  - JOUR
A1  - Maraun, Douglas
A1  - Kurths, Jürgen
T1  - Cross wavelet analysis: significance testing and pitfalls
N2  - In this paper, we present a detailed evaluation of cross wavelet analysis of bivariate time series. We develop a statistical test for zero wavelet coherency based on Monte Carlo simulations. If at least one of the two processes considered is Gaussian white noise, an approximative formula for the critical value can be utilized. In a second part, typical pitfalls of wavelet cross spectra and wavelet coherency are discussed. The wavelet cross spectrum appears to be not suitable for significance testing the interrelation between two processes. Instead, one should rather apply wavelet coherency. Furthermore we investigate problems due to multiple testing. Based on these results, we show that coherency between ENSO and NAO is an artefact for most of the time from 1900 to 1995. However, during a distinct period from around 1920 to 1940, significant coherency between the two phenomena occurs
Y1  - 2004
SN  - 1023-5809
ER  - 
TY  - THES
A1  - Maraun, Douglas
T1  - What can we learn from climate data? : Methods for fluctuation, time/scale and phase analysis
T1  - Was können wir aus Klimadaten lernen? : Methoden zur Fluktuations-, Zeit/Skalen- und Phasenanalyse
N2  - Since Galileo Galilei invented the first thermometer, researchers have tried to understand the complex dynamics of ocean and atmosphere by means of scientific methods. They observe nature and formulate theories about the climate system. Since some decades powerful computers are capable to simulate the past and future evolution of climate. Time series analysis tries to link the observed data to the computer models: Using statistical methods, one estimates characteristic properties of the underlying climatological processes that in turn can enter the models. The quality of an estimation is evaluated by means of error bars and significance testing. On the one hand, such a test should be capable to detect interesting features, i.e. be sensitive. On the other hand, it should be robust and sort out false positive results, i.e. be specific.  This thesis mainly aims to contribute to methodological questions of time series analysis with a focus on sensitivity and specificity and to apply the investigated methods to recent climatological problems.  First, the inference of long-range correlations by means of Detrended Fluctuation Analysis (DFA) is studied. It is argued that power-law scaling of the fluctuation function and thus long-memory may not be assumed a priori but have to be established. This requires to investigate the local slopes of the fluctuation function. The variability characteristic for stochastic processes is accounted for by calculating empirical confidence regions. The comparison of a long-memory with a short-memory model shows that the inference of long-range correlations from a finite amount of data by means of DFA is not specific. When aiming to infer short memory by means of DFA, a local slope larger than $\alpha=0.5$ for large scales does not necessarily imply long-memory. Also, a finite scaling of the autocorrelation function is shifted to larger scales in the fluctuation function. It turns out that long-range correlations cannot be concluded unambiguously from the DFA results for the Prague temperature data set.  In the second part of the thesis, an equivalence class of nonstationary Gaussian stochastic processes is defined in the wavelet domain. These processes are characterized by means of wavelet multipliers and exhibit well defined time dependent spectral properties; they allow one to generate realizations of any nonstationary Gaussian process. The dependency of the realizations on the wavelets used for the generation is studied, bias and variance of the wavelet sample spectrum are calculated. To overcome the difficulties of multiple testing, an areawise significance test is developed and compared to the conventional pointwise test in terms of sensitivity and specificity. Applications to Climatological and Hydrological questions are presented. The thesis at hand mainly aims to contribute to methodological questions of time series analysis and to apply the investigated methods to recent climatological problems.  In the last part, the coupling between El Nino/Southern Oscillation (ENSO) and the Indian Monsoon on inter-annual time scales is studied by means of Hilbert transformation and a curvature defined phase. This method allows one to investigate the relation of two oscillating systems with respect to their phases, independently of their amplitudes. The performance of the technique is evaluated using a toy model. From the data, distinct epochs are identified, especially two intervals of phase coherence, 1886-1908 and 1964-1980, confirming earlier findings from a new point of view. A significance test of high specificity corroborates these results. Also so far unknown periods of coupling invisible to linear methods are detected. These findings suggest that the decreasing correlation during the last decades might be partly inherent to the ENSO/Monsoon system. Finally, a possible interpretation of how volcanic radiative forcing could cause the coupling is outlined.
N2  - Seit der Erfindung des Thermometers durch Galileo Galilei versuchen Forscher mit naturwissenschaftlichen Methoden die komplexen Zusammenhänge in der Atmosphäre und den Ozeanen zu entschlüsseln. Sie beobachten die Natur und stellen Theorien über das Klimasystem auf. Seit wenigen Jahrzehnten werden sie dabei von immer leistungsfähigeren Computern unterstützt, die das Klima der Erdgeschichte und der nahen Zukunft simulieren.  Die Verbindung aus den Beobachtungen und den Modellen versucht die Zeitreihen­analyse herzustellen: Aus den Daten werden mit statistischen Methoden charak­teristische Eigenschaften der zugrundeliegenden klimatologischen Prozesse geschätzt, die dann in die Modelle einfliessen können. Die Bewertung solch einer Schätzung, die stets Messfehlern und Vereinfachungen des Modells unterworfen ist, erfolgt statistisch entweder mittels Konfidenzintervallen oder Signifikanztests. Solche Tests sollen auf der einen Seite charakteristische Eigenschaften in den Daten erkennen können, d.h. sie sollen sensitiv sein. Auf der anderen Seite sollen sie jedoch auch keine Eigenschaften vortäuschen, d.h. sie sollen spezifisch sein. Für die vertrauenswürdige Untermauerung einer Hypothese ist also ein spezifischer Test erforderlich.  Die vorliegende Arbeit untersucht verschiedene Methoden der Zeitreihenanalyse, erweitert sie gegebenenfalls und wendet sie auf typische klimatologische Frage­stellungen an. Besonderes Augenmerk wird dabei auf die Spezifizität der jeweiligen Methode gelegt; die Grenzen möglicher Folgerungen mittels Datenanalyse werden diskutiert. Im ersten Teil der Arbeit wird studiert, wie und ob sich mithilfe der sogenannten trendbereinigenden Fluktuationsanalyse aus Temperaturzeitreihen ein sogenanntes langes Gedächtnis der zugrundeliegenden Prozesse herleiten lässt. Solch ein Gedächtnis bedeutet, dass der Prozess seine Vergangenheit nie vergisst, mit fundamentalen Auswirkungen auf die gesamte statistische Beurteilung des Klimasystems. Diese Arbeit konnte jedoch zeigen, dass die Analysemethode vollkommen unspezifisch ist und die Hypothese “Langes Gedächtnis” gar nicht abgelehnt werden kann.  Im zweiten Teil werden zunächst Mängel einer sehr populären Analysemethode, der sogenannten kontinuierlichen Waveletspetralanalyse diskutiert. Diese Methode schätzt die Variabilität eines Prozesses auf verschiedenen Schwingungsperioden zu bestimm­ten Zeiten. Ein wichtiger Nachteil der bisherigen Methodik sind auch hier unspezi­fische Signifikanztests. Ausgehend von der Diskussion wird eine Theorie der Wavelet­spektralanalyse entwickelt, die ein breites Feld an neuen Anwendungen öffnet. Darauf basierend werden spezifische Signifikanztests konstruiert. Im letzten Teil der Arbeit wird der Einfluss des El Niño/Southern Oscillation Phäno­mens auf den Indischen Sommermonsun analysiert. Es wird untersucht, ob und wann die Oszillationen beider Phänomene synchron ablaufen. Dazu wird eine etablierte Methode für die speziellen Bedürfnisse der Analyse von typischerweise sehr unregel­mäßigen Klimadaten erweitert. Mittels eines spezifischen Signifikanztests konnten bisherige Ergebnisse mit erhöhter Genauigkeit bestätigt werden. Zusätzlich konnte diese Methode jedoch auch neue Kopplungsintervalle feststellen, die die Hypothese entkräften konnten, dass ein neuerliches Verschwinden der Kopplung ein beisspielloser Vorgang sei. Schliesslich wird eine Hypothese vorgestellt, wie vulkanische Aerosole die Kopplung beeinflussen könnten.
KW  - Spektralanalyse <Stochastik>
KW  - Monsun
KW  - Klimatologie
KW  - Zeitreihenanalyse
KW  - Wavelet-Analyse
KW  - El-Niño-Phänomen
KW  - Kopplungs-Analyse
KW  - Kohärenz-Analyse
KW  - Phasen-Analyse
KW  - Signifikanztests
KW  - Continuous Wavelet Spectral Analysis
KW  - Wavelet Coherence
KW  - Phase-Analysis
KW  - Significance Testing
KW  - Climatology
Y1  - 2006
U6  - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:kobv:517-opus-9047
ER  -