Causes of Agricultural Drought

Interannual Rainfall Variability

Southern Africa is a predominantly semiarid region with high interannual variability in rainfall (figure 20.4) and a pronounced annual cycle (Nicol-son, 1986; Tyson, 1986). The coefficient of variation of annual rainfall

Figure 20.2 Time series of Zimbabwe smallholder-farming sector maize yields and annual rainfall from 1970 to 1999.

over Zimbabwe, for example, exceeds 40% in the drier west and southwest sections of the country and is below 25% in the north (figure 20.5). Recent studies of interannual rainfall variability over southern Africa have demonstrated their periodic nature. Historical rainfall records generally show spectral peaks in five bands, 2.2-2.4, 2.6-2.8, 3.3-3.8, 5-7, and 1720 years (Nicholson, 1986; Makarau and Jury, 1997). It has been argued that these periodicities in annual rainfall are indicative of the influence of the Quasi-Biennial Oscillation, ENSO, periodic sea-surface temperature oscillations, and the luni-solar cycles (Nicholson, 1986; Makarau and Jury, 1997).


Considerable evidence exists that phase shifts in the ENSO are accompanied by rainfall anomalies across southern Africa (Ropelewski and Halpert,

Figure 20.3 Response of Zimbabwe's GDP growth rate to rainfall.


Figure 20.4 Zimbabwe's annual rainfall departures from the long-term mean from 1901 to 2000. The smooth curve represents the 10-year moving average.


Figure 20.4 Zimbabwe's annual rainfall departures from the long-term mean from 1901 to 2000. The smooth curve represents the 10-year moving average.

1987; Matarira, 1990; Cane et al., 1994). A detailed review of ENSO can be found in chapter 3. It has been argued that the influence of ENSO events is strongest during the peak austral summer rainfall months of December-March because that is when the warm and cold events have reached maturity and when the upper westerlies have retreated significantly poleward (Mason, 2001). This delayed rainfall response has a potential value in operational long-range forecasting (Cane et al., 1994). Rainfall is reduced by 20-60% during some warm ENSO events across Zimbabwe for both the first and second part of the rainfall season. Rainfall deficits tend to be greatest in the southeast section of the country. Using maize yields from Zimbabwe's smallholder-farming sector, Cane et al. (1994) showed that more than 60% of the observed variation in yield could be predicted from Niño 3 sea-surface temperature anomalies several months in advance.

Sea-Surface Temperatures

Several studies have shown that summer rainfall in southern Africa responds to anomalous global sea-surface temperature (SST) changes (Cane et al., 1994; Makarau and Jury, 1997; Rocha and Simmonds, 1997). A warmer or cooler than normal eastern equatorial Pacific tends to be associated with dry or wet conditions, respectively, across the country (Cane et al., 1994). This response is part of the well-documented ENSO cycle (Ro-pelewski and Halpert, 1987). It has been shown that anomalously warm SSTs in the central equatorial Indian Ocean are usually associated with dry conditions over the central and southern parts of southern Africa. The SST anomaly pattern of the Indian Ocean could be important in the transmission of the El Niño signal to southern Africa (Rocha and Simmonds, 1997). However, it has also been observed that occasionally the Indian Ocean warm events can occur independent of ENSO events (Mason, 2001).

Figure 20.5 Coefficient of variation (%) of Zimbabwe's annual rainfall (1961-2000).

Associations between southern African interannual rainfall fluctuations and SSTs in the Indian Ocean appear to be complex. Although dry conditions are frequently associated with a warmer than normal western tropical Indian Ocean, this area is also an important source of atmospheric moisture for southern Africa. It becomes the dominant source during the second half of summer (Rocha and Simmonds, 1997), implying that an increase in SST here could enhance rainfall over most parts of southern Africa. A warmer tropical Indian Ocean also enhances the chances of tropical cyclone formation. Depending on the track, the tropical cyclones forming in the tropical Indian Ocean may act to dry out eastern sections of the subregion or bring floods. With a warm tropical Indian Ocean being capable of producing two opposite climatic effects, SST-based seasonal climate forecasting becomes a complex operation for much of southern Africa.

0 0

Post a comment