This dissertation assesses the extent to which data from the Climatological Database for the World's Oceans (CLIWOC) reflect newly digitized historical wind data captured at the Royal Astronomical Observatory (RAO) in Cape Town, South Africa from 1834-1854. This follows the historical precipitation reconstructions for Southern Africa by Hannaford et al. (2015), using wind data from the CLIWOC database. This project also forms part of a bigger project that is recovering and digitising historical instrumental meteorological data for Southern Africa that have never been analysed before. For Southern Africa, the opportunity to compare historical instrumental data seldom arises due to the paucity of reliable data. However, there is an opportunity to analyse and compare two different wind data sources for a twenty-one year cross over period for south western Africa. Wind, as an indicator of atmospheric conditions, has not been assessed extensively in South African, therefore this project fills an academic gap in historical climatology for the region, and provides newly digitised historical data. Digitisation and pre-processing steps ensure that the RAO dataset is comparable to the CLIWOC dataset. This is done by replicating wind direction and speed measurement conversions and formatting (Garcia-Herrera et al., 2005), and by mirroring the available time steps of data in each dataset (eliminating data were the other dataset has erroneous or missing data). Spatially scattered data recorded over the sea compared to data recorded at a fixed position introduces inherent limitations, error and noise into the data comparison. Therefore, to eliminate as many uncertainties as possible and minimise the noise in the data, the CLIWOC data are refined further by a) a single observation per day, b) separating three regions of differing seasonal synoptic air flow regimes (west coast, south west peninsula and south coast) and c) all analyses based on seasonally grouped data. Temporal, spatial and vector relationships are established for each season using scatter plot graphs and Pearson correlations. The different relationships between the data are derived from corresponding wind data (i.e. data of the same day and time), in each dataset for wind speed and wind direction separately. No significant correlation (all p values>0.05) or signal is evident over time, or as the difference in distance changes. However, seasonality is represented consistently in the wind vector distribution heat maps. Significant findings include the observations of anomalous north westerly winds in summer at the RAO, where the CLIWOC data did not pick up similar data for the corresponding region on the west coast. Historical wind data used herein prove to be reliable by the expected seasonal synoptic flow patterns and characteristics seen in each study region. There is no correlation between the datasets over time and space and the data do not present any clear signals or return events over time. Although corresponding data do not show any correlations, there are typical synoptic flow regimes in each study region which prove that wind data was recorded correctly. Therefore, the datasets are mutually exclusive, but accurate in their intrinsic value. It is only the anomalous summer north westerlies at the RAO which question the reliability of the data, as the same wind regimes were not identifiable in the corresponding CLIWOC data. This anomaly was noted but not studied further. This project highlights the major inconsistencies and limitations in the CLIWOC data. Researchers in the future should use CLIWOC data appropriately to suit the research question and be aware of the inconsistencies that may introduce noise.
| Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:uct/oai:localhost:11427/24453 |
| Date | January 2017 |
| Creators | Brown, Alexa |
| Contributors | Lennard, Chris, Grab, Stefan |
| Publisher | University of Cape Town, Faculty of Science, Department of Environmental and Geographical Science |
| Source Sets | South African National ETD Portal |
| Language | English |
| Detected Language | English |
| Type | Master Thesis, Masters, MSc |
| Format | application/pdf |
Page generated in 0.0021 seconds