Quantifying biases and exploring natural climate variability in early instrumental temperature records

Wallis, Emily Jayne (2024) Quantifying biases and exploring natural climate variability in early instrumental temperature records. Doctoral thesis, University of East Anglia.

[thumbnail of EW 242524 Final PhD Thesis.pdf]
Download (14MB) | Preview


Early instrumental land surface air temperature (LSAT) records are valuable resources for climate science; however, many are compromised by the presence of non-climatic influences (inhomogeneities) which must be addressed before records can be used for climate analysis. This thesis presents an empirical approach to address a pervasive inhomogeneity affecting LSAT records – exposure bias – before using the exposure-bias-adjusted data, two reanalyses and model simulations to explore the climatic influence of volcanic eruptions; specifically, whether there is early instrumental evidence that explosive volcanism causes post-eruption Eurasian winter warming.

Exposure biases occur due to changes in the way thermometers have been exposed to the elements over time and have not been widely accounted for in global LSAT compilations. This thesis addresses the exposure bias in an extended version of CRUTEM5 (starting in 1781) by (1) analysing 54 parallel measurement series to better characterise the bias; (2) developing three statistical models to predict the bias from temperature and radiation variables; and (3) applying the models to mid-latitude stations in CRUTEM5 using a compilation of exposure metadata collated here. The models are applied to 1,960 stations, cooling summer temperatures in the Northern Hemisphere mid-latitudinal mean by ~0.2°C between 1882–1934, with smaller adjustments in other seasons and time periods.

Analysis of the first and second winters following eight explosive volcanic eruptions finds a consistent and significant volcanic signal of warmer winter temperatures in Europe, cooler near Greenland and a tendency for a positive NAO index in both the observations and reanalyses. In contrast, the UKESM1.1 model simulates a similar response only following the largest eruption, Tambora. The consistency of the volcanic signal in the observations and reanalyses, plus the modelled response following Tambora, suggests explosive volcanism does contribute to observed post-eruption Eurasian winter warming and the eruptions in the early instrumental period strengthen this evidence.

Item Type: Thesis (Doctoral)
Faculty \ School: Faculty of Science > School of Environmental Sciences
Depositing User: Nicola Veasy
Date Deposited: 10 Jul 2024 10:44
Last Modified: 10 Jul 2024 10:44
URI: https://ueaeprints.uea.ac.uk/id/eprint/95854


Downloads per month over past year

Actions (login required)

View Item View Item