Spatio-temporal Statistics for the Transition of Energy and Transport
Growing digitalization creates new challenges for statistical data analysis due to the increasing availability of various kinds of massive data, which are nowadays often automatically collected at many different locations with possibly different sampling frequencies over time. New statistical methodology is needed for parsimonious modeling and proper analysis of these resources, in order to deduce the relevant information from it and to create empirically validated knowledge. This is crucial for many tasks such as workflow optimization, prediction of future evolutions, assessment of inherent uncertainties and the early detection of disturbances. It applies in particular to the production and transport of energy and goods, which have to be designed as efficient as possible with respect to the consumption of financial, temporal and also of natural resources. Technological advances and new policies need to promote welfare of mankind while respecting the integrity of the ecosystem without adversely affecting living standards. Overall, the reduction of CO2 emissions and the transition to a low-carbon economy and renewable energy are among the most important global challenges for the next decades. To surmount these challenges, we need to base decision-making on reliable and comprehensive information such that its impact on whole economies, multi-national bodies, and people's everyday life can be accurately predicted and accompanied by statements about the uncertainties of such predictions. The objective of the planned Collaborative Research Center/Transregio (TRR) is the development of novel and innovative statistical methodologies to extract reliable information from spatio-temporal data in order to support data-based decision-making in technological, economic, and ecological applications. More specifically, we aim to develop tailor-made methods for solving challenging data analytical problems in the areas of statistics for the transport of energy and goods and econometrics of the energy and mobility transition. The foundation of our TRR is an area devoted to general statistical methodologies for spatio-temporal data, which, in addition to several content-related links, establishes the essential connections between the two application areas. Here – motivated by specific applications – common statistical problems will be identified and fundamental methodology for statistical inference will be developed and investigated. Statistical and data-analytical solutions of concrete problems will be provided in close cooperation with the application-driven projects. By this interdisciplinary approach, we will construct new statistical inference tools for spatio-temporal data, which will be broadly adaptable also in other fields of applications. Our research program builds on the particular strengths of TU Dortmund University in statistics, econometrics, logistics and electrical engineering, and of Ruhr University Bochum in mathematical statistics, deep learning, econometrics and energy economics. These shall be combined and complemented with specific expertises from neighboring institutions, such as renewable energy and econometrics at the University of Duisburg-Essen and RWI Essen, and spatio-temporal modeling at the University of Münster. The team is completed with a project leader from Humboldt-Universität zu Berlin, with expertise on modeling spatio-temporal dependencies.
Antragsteller
Prof. Dr. Roland Fried, TU Dortmund, Fakultät Statistik
Förderlinie: Sprint
Gesamtfördersumme: 24.729,60 €
Ansprechpartner
Prof. Dr. Roland Fried
TU Dortmund
Fakultät Statistik
Vogelpothsweg 87
44221 Dortmund
Email: roland.fried@tu-dortmund.de