Agricultural perspectives on the 2015-2018 Western Cape drought, South Africa: Characteristics and spatial variability in the core wheat growing regions.

• Places the 2015-2017 drought in both a historical and future context. • Wheat regions experienced persistent drought with high spatial-temporal variability. • The drought was severe and could be an early sign of climate change in the area. Between 2015 and 2018, the Western Cape Province of South Africa experienced a multi-year severe drought, which negatively impacted major economic sectors. The province serves as an important producer of wheat in South Africa among other agricultural commodities. This study aims to analyze the 2015-2018 drought and its impacts on wheat production in the winter rainfall region of the Western Cape, South Africa. The central focus is to place the drought in both a historical and future context while emphasising the differences between the two core wheat growing regions. We present an analysis of the drought, as well as climate trends computed from weather data sets (1988–2018) from six weather stations across the two core wheat-growing. We first present a meteorological drought analysis of mean annual and seasonal rainfall and temperatures, subsequently providing an agricultural lens by computing Standardised Precipitation and Evapotranspiration Index (SPEI) accumulated over 12- and 36-month timescales, self-calibrated Palmer Drought Severity Index (sc-PDSI), changes to the start and end of the rainfall season, and their effects on wheat yields. Trend analysis was conducted to determine if drought observations form part of the longer-term trends in the region. Finally, we show how the drought varied spatially across the two regions. Results show that between 1988 and 2018, the wheat growing areas of the Western Cape experienced persistent drought with high spatial-temporal variability. The 2015-2018 drought, however, was the most severe experienced in the 30-year study period at five of the six stations. These results are consistent with conditions that can be expected under future climate change. Moreover, results can be useful for the development of early warning systems since they place the drought in the context of past drought conditions. [ABSTRACT FROM AUTHOR]