• Refine Query
  • Source
  • Publication year
  • to
  • Language
  • 10
  • 3
  • 2
  • 1
  • Tagged with
  • 22
  • 22
  • 13
  • 8
  • 4
  • 4
  • 3
  • 3
  • 3
  • 3
  • 3
  • 3
  • 3
  • 2
  • 2
  • About
  • The Global ETD Search service is a free service for researchers to find electronic theses and dissertations. This service is provided by the Networked Digital Library of Theses and Dissertations.
    Our metadata is collected from universities around the world. If you manage a university/consortium/country archive and want to be added, details can be found on the NDLTD website.
21

Use of large-scale atmospheric flow patterns to improve forecasting of extreme precipitation in the Mediterranean region for longer-range forecasts

Mastrantonas, Nikolaos 31 May 2023 (has links)
The Mediterranean region frequently experiences extreme precipitation events (EPEs) with devastating consequences for affected societies, economies, and environment. Thus, it is crucial to better understand their characteristics and drivers and improve their predictions at longer lead times. This work provides new insights about the spatiotemporal dependencies of EPEs in the region. It, moreover, implements Empirical Orthogonal Function analysis and subsequent non-hierarchical Kmeans clustering for generating nine distinct weather patterns over the domain, referred to as “Mediterranean patterns”. These patterns are significantly associated with EPEs across the region, and in fact, can be used to extend the forecasting horizon of EPEs. This is demonstrated considering modelled data for all the domain, but also using observational data for Calabria, southern Italy, an area of complex topography that increases the challenges of weather prediction. The results suggest preferential techniques for improving EPEs predictions for short, medium, and extended range forecasts, supporting thus the mitigation of their negative impacts.
22

Complex network analysis of extreme rainfall in South America

Boers, Niklas 01 June 2015 (has links)
Basierend auf der Theorie von Netzwerken wird ein allgemeines Rahmenwerk entwickelt, um kollektive Synchronisationsphänome von Extremereignissen in komplexen Systemen zu studieren. Die Methode vergleicht die Variabilität der einzelnen Teile des Systems auf Grundlage von Beobachtungszeitreihen mit dem Ziel, emergente Synchronisationsmuster von Extremereignissen auf makroskopischer Ebene aufzudecken. Zu diesem Zweck werden die einzelnen Zeitreihen eines interaktiven Systems mit den Knoten eines Netzwerks identifiziert und die Abhängigkeiten zwischen diesen durch die Kanten des Netzwerks dargestellt. Die komplexe interne Synchronisationsstruktur des Systems wird so in Form der Netzwerktopologie mathematisch zugänglich gemacht und kann durch die Einführung geeigneter Netzwerkmaße analysiert werden. Die Methode wird im Folgenden auf räumlich und zeitlich hochaufgelöste Regendaten aus Satellitenmessungen angewendet, um die kollektive Dynamik extremer Regenereignisse in Südamerika zu untersuchen. Diese Anwendung verfolgt drei Ziele: Erstens wird gezeigt, wie die hier entwickelte Methode zur klimatologischen Analyse verwendet werden kann. Zweitens können Quellen und Senken von Extremereignissen durch die Einführung des Konzeptes der Netzwerkdivergenz identifiziert werden. Dies erlaubt es, die gerichteten Netzwerkpfade, entlang derer Extremereignisse synchronisieren, nachzuverfolgen. Auf dieser Grundlage wird eine statistische Regel gewonnen, die beträchtliche Anteile der extremen Regenereignisse in den Zentralanden vorhersagt. Drittens werden die bis dahin entwickelten Methoden und gewonnenen Einsichten dazu verwendet, die Darstellung extre- mer Regenereignisse in verschiedenen Datensätzen zu vergleichen. Insbesondere wird in diesem Kontext die Implementierung solcher Ereignisse in drei gängigen Klimamodellen evaluiert. / Based on the theory of networks, a general framework is developed to study collective synchronization phenomena of extreme events in complex systems. The method relies on observational time series encoding the variability of the single parts of the system, and is intended to reveal emerging patterns of extreme event synchronization on the macroscopic level. For this purpose, the time series obtained from an interactive system under consideration are identified with network nodes, and the possibly delayed and non-linear interdependence of extreme events in different time series is represented by network links connecting the nodes. In this way, the complex internal synchronization structure of the system becomes accessible in terms of the topology of the network, which can be analyzed by introducing suitable network measures. The methodology is applied to satellite-derived rainfall time series of high spatiotemporal resolution in order to investigate the collective dynamics of extreme rainfall events in South America. The purpose of this application is threefold: First, it is shown how the methodology can be used for climatic analysis by revealing climatological mechanism from the spatial patterns exhibited by different network measures. Second, by introducing the concept of network divergence, sink and source regions of extreme events can be identified, allowing to track their directed synchronization pathways through the network. A simple statistical forecast rule is derived on this basis, predicting substantial fractions of extreme rainfall events in the Central Andes. Third, the methodology and the insights developed in the first two steps are used to evaluate the dynamical representation of extreme events in different datasets, and in particular their dynamical implementation in three state of the art climate models.

Page generated in 0.0775 seconds