A paper emerging from the research of agrometeorology graduate Dr Floyd Khosa, now a data scientist at Santam Insurance, received the South African Society for Atmospheric Sciences (SASAS) Stanley Jackson Award for the best research paper in atmospheric sciences in South Africa in 2020.
Named after the late Professor Stanley Jackson – an eminent meteorologist who established the University of the Witwatersrand’s school of climatology and meteorology, helped found SASAS, and focused his pioneering research on the circulation of the atmosphere over Southern Africa – the award recognises notable contributions to the atmospheric and oceanic sciences in South Africa.
The accolade is presented for the best peer review paper published within the two years before the conference.
The result of research done for Khosa’s PhD at UKZN under the supervision of Professor Mike Savage, the paper is titled: Evaluation of Modelled Actual Evapotranspiration Estimates from a Land Surface, Empirical and Satellite-Based Models Using in situ Observations from a South African Semi-Arid Savanna Ecosystem. It was published in the Agricultural and Forest Meteorology Journal while Khosa was a candidate researcher at the Council for Scientific and Industrial Research (CSIR).
Khosa graduated in April 2020 after completing his thesis on the topic of multi-model estimates of evapotranspiration and soil moisture in the context of climate change across South African landscapes.
This paper dealt with models used to estimate evapotranspiration (ET) – the transfer of water vapour from a surface into the atmosphere. Khosa explained that ET played a vital role in the land-atmosphere interaction and in climate variability, particularly in arid and semi-arid areas. Precisely estimating ET is important in hydrological, water resource and climate modelling, so Khosa sought to evaluate the accuracy of eight ET data products derived from models that are used to achieve these estimates.
Khosa’s research highlights that methods of estimating ET are inherently limited by targeting specific temporal and spatial scales, and by their own assumptions, errors and technical challenges. Tackling the limitations of ET modelling that are common worldwide, but even more constrained for semi-arid regions, Khosa made inroads into evaluating ET models for semi-arid ecosystems. His aim was to establish if the models can be used reliably to simulate local and regional ET conditions in Africa, comparing the models’ ET estimates with in situ data acquired from a flux tower in the Kruger National Park.
He found that all models assessed overestimate ET during summer and underestimate it in winter, and concluded that satellite-derived model ET outputs have the potential to aid understanding of ET’s spatiotemporal variability across different landscapes, and said that process-based models could be used for climate change impact studies on ET. He also suggested that future studies could assess the models within even more African biomes and savanna types to evaluate their accuracy.
The paper was co-authored with Dr Gregor Feig of the South African Environmental Observation Network (SAEON), Mrs Martina van der Merwe, Dr Mohau Mateyisi and Mr Azwitamisi Mudau of the CSIR, and Savage.
Words: Christine Cuénod