Abstract
Seasonal-to-interannual variations of rainfall over southern Africa, key to predicting extreme climatic events, are predictable over certain regions and during specific periods of the year. This predictability had been established by testing seasonal forecasts from models of varying complexity against official station rainfall records typically managed by weather services, as well as against gridded data sets compiled through a range of efforts. Members of the general public, including farmers, additionally have extended records of rainfall data, often as daily values spanning several decades, which are recorded and updated regularly at their farms and properties. In this paper, we show how seasonal forecast modelers may use site recorded farm rainfall records for the development of skillful forecast systems specific to the farm. Although the uptake of seasonal forecasts in areas with modest predictability such as southern Africa may be challenging, we will show that there is potential for financial gain and improved disaster risk farm management by co-developing with farmers forecast systems based on a combination of state-of-the-art climate models and farm rainfall data. This study investigates the predictability of seasonal rainfall extremes at five commercial farms in southern Africa, four of which are in the austral summer rainfall areas, while one is located in the winter rainfall area of the southwestern Cape. We furthermore calculate a measure of cumulative profits at each farm, assuming a “fair odds” return on investments made according to forecast probabilities. The farmers are presented with hindcasts (re-forecasts) at their farms, and potential financial implications if the hindcasts were used in decision-making. They subsequently described how they would use forecasts for their farm, based on their own data.
Highlights
Seasonal forecast model development has a long history in South Africa (Landman, 2014)
Statistical post-processing is required to improve on seasonal forecasts from global models
Statistical correction still forms part of the seasonal forecast systems developed at certain institutions for operational forecasting
Summary
Seasonal forecast model development has a long history in South Africa (Landman, 2014). Notwithstanding the proven accuracies of seasonal climate forecast models, statistical correction methods are recommended even for today’s coupled climate model forecasts (Barnston and Tippett, 2017) Such statistical correction has had a long track record in terms of testing predictability over southern African countries including South Africa and Namibia (e.g., Bartman et al, 2003; Landman and Goddard, 2005), where local climatic mechanisms effecting seasonal-to-interannual variations over these countries have been studied comprehensively (e.g., Tadross et al, 2005; Hansingo and Reason, 2009; Reason and Smart, 2015). Such a process of co-learning between forecast modelers and forecast users may help improve on seasonal forecasts tailored for commercial farm management, and, potentially improve forecast uptake—a challenging aspect in southern Africa, where seasonal forecasting skill ranks modestly with other regions globally (Landman et al, 2019)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
