Seasonal rainfall variability in southeast Africa during the nineteenth century reconstructed from documentary sources

Abstract

Analyses of historical patterns of rainfall variability are essential for understanding long-term changes in precipitation timing and distribution. Focussing on former Natal and Zululand (now KwaZulu-Natal, South Africa), this study presents the first combined annual and seasonal reconstruction of rainfall variability over southeast Africa for the 19th century. Analyses of documentary sources, including newspapers and colonial and missionary materials, indicate that the region was affected by severe or multi-year drought on eight occasions between 1836 and 1900 (the rainy seasons of 1836-38, 1861-63, 1865-66, 1868-70, 1876-79, 1883-85, 1886-90 and 1895-1900). Six severe or multi-year wet periods are also identified (1847-49, 1854-57, 1863-65, 1879-81, 1890-91 and 1892-94). The timing of these events agrees well with independent reconstructions of 19th century rainfall for other parts of the southern African summer rainfall zone (SRZ), suggesting subcontinental scale variability. Our results indicate that the relationship between El Niño and rainfall in southeast Africa was relatively stable, at least for the latter half of the 19th century. El Niño conditions appear to have had a more consistent modulating effect upon rainfall during the 19th century than La Niña. The rainfall chronology from this study is combined with other annually-resolved palaeoclimate records from mainland southern Africa and surrounding oceans as part of a multi-proxy rainfall reconstruction for the SRZ. This reconstruction confirms (i) the long-term importance of ENSO and Indian Ocean SSTs for modulating regional rainfall; and (ii) that summer precipitation has been declining progressively over the last 200 years

African hydroclimatic variability during the last 2000 years

“…Nicholson, 1981;Vogel, 1989;Nicholson, 2000;Endfield, 2002a, 2002b;Kelso and Vogel, 2007;Nash and Endfield, 2008;Nash and Grab, 2010;Norrgård, 2015;Nash et al, 2016), tree-ring records (e.g. Dunwiddie and LaMarche, 1980;Therrell et al, 2006;Woodborne et al, 2015), reconstructions from ships' logbooks (Hannaford et al, 2015) and, from a handful of countries, early rain-gauge data (Nicholson et al, 2012a;Nicholson et al, 2012b).…”

Back-to-the-future potential for autochthonous transmission of Aedes aegypti-transmitted viruses in eThekwini and urban coastal KwaZulu-Natal Province, South Africa

“…[27][28] Eight of those 12 years saw rainfall in Durban below the historical average (Table 1), often coincident with El Niño-associated drought. [29][30][31] While the precise viral aetiologies of outbreaks and epidemics prior to 1926/27 are elusive, the better documented occurrences of probable chikungunya or dengue point to the presence of human-biting, virus-transmitting Ae. aegypti populations and recurring intersections with water stress.…”

“…Aedes aegypti-transmitted viruses in eThekwini and urban coastal KwaZulu-Natal Province, South Africa [14] Recorded rainfall [14] (mm)* Rainfall as % more or less than annual average 1871 -1997 (1 002 mm) [31] 1873 EDITORIAL may still collect and store water to hedge against prohibitive costs or restrictions and interruptions. This is not to suggest that eThekwini's well-to-do are unlikely to be affected.…”

Comparing proxy and model estimates of hydroclimate variability and change over the Common Era

“…China, Japan, and Europe comprise the three main geographic regions where an abundance of historical climatological sources exist. More recently, assessments have been initiated for records from South America that date back to the beginning of Spanish colonization (Neukom et al, 2010). Data from Africa, North America, and Australia are available for much shorter periods into the past in comparison with other regions where more abundant documentary data exist (e.g., Grab and Nash, 2010;Nash and Grab, 2010;Fenby and Gergis, 2013;Gergis and Ashcroft, 2013;Neukom et al, 2014;Nash et al, 2016).…”

“…Initial efforts to compare proxies and models over the CE largely focused on temperature (e.g., Mann et al, 2009;Widmann et al, 2010;Goosse et al, 2012a;Fernandez-Donado et al, 2013;Masson-Delmotte et al, 2013;Phipps et al, 2013;Lehner et al, 2015;PAGES 2k-PMIP3 Group, 2015;Luterbacher et al, 2016), but more recent efforts have considered aspects of hydroclimate (e.g., Coats et al, 2013bCoats et al, , 2015aLandrum et al, 2013;Tierney et al, 2013;Gómez-Navarro et al, 2015;Smerdon et al, 2015;Ljungqvist et al, 2016;OttoBliesner et al, 2016;Stevenson et al, 2016;Li et al, 2017). These hydroclimate proxy-model comparisons have been made possible not only by the growing ensemble of lastmillennium simulations, but also by the expanding number of hydroclimate proxies and their syntheses into largescale gridded hydroclimate reconstructions (e.g., Cook et al, 2004Cook et al, , 2010aCook et al, , 2015bPauling et al, 2006;Jones et al, 2009;Neukom et al, 2010;Masson-Delmotte et al, 2013;Palmer et al, 2015;Brewer et al, 2007;Carro-Calvo et al, 2013).…”