Reconstruction of the historical climate of the Southern Ocean from whaling ships’ logbooks
The estimated century-long trends in the Southern Ocean climate are uncertain, as the data used to estimate these changes are relatively short (~60 years) compared to possible long-term climate variability found in the region. Due to the lack of longer land-based meteorological records, logbooks of...
| Main Author: | |
|---|---|
| Format: | Doctoral or Postdoctoral Thesis |
| Language: | English |
| Published: |
University of Cambridge
2020
|
| Subjects: | |
| Online Access: | https://www.repository.cam.ac.uk/handle/1810/308474 https://doi.org/10.17863/CAM.55560 |
| Summary: | The estimated century-long trends in the Southern Ocean climate are uncertain, as the data used to estimate these changes are relatively short (~60 years) compared to possible long-term climate variability found in the region. Due to the lack of longer land-based meteorological records, logbooks of Christian Salvesen Whaling Company’s whaling ships operating in the Southern Ocean in the 1930s and 1950s are investigated in this thesis. A historical climate dataset is produced from the meteorological observations from the whaling logbooks. The mean sea-level pressure (MSLP) changes across the Southern Ocean diverge: for example, in the northern reaches (55°S latitudinal band) MSLP in the historical period (the 1930s and 1950s) is found to be lower than modern climatology (1981-2010). It is in contrast to the southern reaches (65°S latitudinal band) where MSLP is found to be decreasing over the same period. A historical Southern Annular Mode (SAM) index from 1930-1960 is generated from the whaling dataset, and a significant positive trend is found. Subsequently, historical cyclonic frequency is estimated using a semi-supervised cyclone identification algorithm. The average number of cyclones per year for the historical period is statistically lower than the average for the 1999-2008 period over a control area in the Weddell Sea. Finally, the whaling dataset is assimilated into two (CERA-20C and 20CRv2c) current- generation reanalyses using an offline data assimilation method. The uncertainty in MSLP fields over the assimilation window decreased in both reanalyses by ∼40% in the area of observations post-assimilation. Overall it is shown that meteorological observations from the whaling logbooks can be utilised to reconstruct historical climate in terms of MSLP variability, climate modes (e.g. SAM) and identification of individual cyclones, and to improve the representation of past climate in the reanalyses. In summary, it has been demonstrated that the Southern Ocean climate near the Antarctic coast in the ... |
|---|