Structure and Causes of Variability in Earth’s Surface Air Temperature at the Local to Global Level

• Main Author: Ellerhoff, Beatrice Marie
• Other Authors: Rehfeld, Kira (Prof. Dr.)
• Format: Doctoral or Postdoctoral Thesis
• Language: English
• Published: Universität Tübingen 2024
• Subjects: Klima; Statistik; Temperatur; Modellierung; Data Science; Umweltphysik; 500; Klimavariabilität; climate variability; climate dynamics; climate statistics; climate modelling; Sea ice
• Online Access: http://hdl.handle.net/10900/142580; https://doi.org/10.15496/publikation-83926; http://nbn-resolving.de/urn:nbn:de:bsz:21-dspace-1425807

Fluctuations in Earth’s surface air temperature determine the climatic conditions for life on our planet and influence, among other factors, the occurrence of extreme climatic events. Therefore, a precise understanding of this variability in temperature is paramount to society, politics, and the economy. However, many characteristics and impacts of temperature variability are still uncertain, especially at the local spatial scale and on decadal to multicentennial timescales. Moreover, the relative contribution of climate system-inherent and external drivers of variability needs to be better constrained. Finally, how temperature variability changes with the mean climate state and, thus, under current and future anthropogenic warming remains to be clarified. These uncertainties affect long-term planning, for example, concerning mitigation and adaptation strategies. In three publications, this dissertation examines the statistical properties of temperature variability as a function of the underlying spatiotemporal scales, external drivers, and mean climate state. Spectral methods are combined with time series analysis, conceptual modeling, and Bayesian inference to quantify temperature variability from climate model simulations and paleoclimate records. The results confirm overall confidence in the simulated global temperature variability. Climate models, however, respond more strongly to external forcing and show fewer internal fluctuations at decadal scales than paleoclimate reconstructions of global temperature. At the local level, there are significant data-model mismatches in temperature variance and correlation properties over decadal and longer timescales. Improved representation of natural forcing in climate model simulations can partially offset these differences. By integrating the response to volcanic eruptions, sea ice dynamics plays an essential role in amplifying local multidecadal variability. Decreasing sea ice extent attenuates local variability under warming, especially in high latitudes. Overall, ...