The following HTML text is provided to enhance online
readability. Many aspects of typography translate only awkwardly to HTML.
Please use the page image
as the authoritative form to ensure accuracy.
Natural Climate Variability on Decade-to-Century Time Scales
Water-level fluctuations of Lake Abiyata, A.D. 1984-1991, based on a time series of AVHRR imagery, together with sample images illustrating the appearance of the lake (center) in mid-1986 and late 1988 (From Harris et al., 1992; reprinted with permission of the European Space Agency.)
Measured water-level fluctuations of Lake Titicaca, A.D. 1912-1970. (Redrawn from Kessler, 1974; reprinted with permission of Dr. Kessler.)
Measured water-level fluctuations of Lake Victoria, A.D. 1899-1974; (Redrawn from Vincent et al., 1979; reprinted with permission of Kluwer Academic Publishers.)
nificant spectral peaks with periods of 10.6 and 2.4 years. The former was assigned to the sunspot cycle and the latter to the quasi-biennial cycle. They also found a very weak period of about 4.7 years, which they ascribed to El Niño. More recent work suggests that both the quasi-biennial 2.4-year and lower-frequency 4.2-year cycles are associated with ENSO and the related variations in atmospheric angular momentum (Keppenne and Ghil, 1992; Dickey et al., 1992).
The Influence of Sea Surface Temperature Anomalies in the Indian Ocean
A lake that at least in part acts as an indicator of sea surface temperature (SST) anomalies in the Indian Ocean is Lake Victoria (1°S) in equatorial Africa. This is a very large, open lake (67,000 km2), with an equilibrium response time te of about 4.5 years (from data in Institute of Hydrology, 1985). Its water level curve for 1899 to 1978 shows the following features: oscillations about a stationary mean during the period 1899 to 1961; a step increase in lake level in 1961-1962; and an oscillating decline over the following 16 years (Figure 4) (Mörth, 1967; Vincent et al., 1979; Institute of Hydrology, 1985).
Spectral analysis of the lake-level record for 1899 to 1974 by Vincent et al. (1979) revealed only two significant peaks, at 10-13 years and 5-6 years, the first of which was attributed to solar variability in a famous paper by Brooks (1923). The record contains no power in the 2-to-3-year range; nor does it show any correlation with the Southern Oscillation Index (Vincent et al., 1979).
The sharp increase in lake level in 1961-1962 is an example of a ''spike" (Figure 4), although a preliminary inversion of the lake-level curve suggests that the climate over the Victoria catchment has not fully returned to its pre-1961 state (I. M. Mason, pers. commun.). The sudden rise in water level was the result of a remarkable rainfall anomaly over East Africa in August 1961 to January 1962, covering approximately 2.1 × 106 km2. The peak occurred in November 1961 when rainfall over the Lake Victoria basin was almost 400 percent of normal (Flohn, 1987). Flohn has summarized evidence suggesting that this was merely the western end of a large, positive anomaly of SST, cloud, and rainfall covering about 6 × 106 km2 of the western Indian Ocean, which is an extent similar to that of a typical ENSO warm pool or of the South Atlantic warming event of 1984 (Philander, 1986).
The Influence of the Atlantic (and Global) Sea Surface Temperatures
Pronounced decade-to-century-scale fluctuations in water balance have affected Lake Chad, situated at 13°N in the West African Sahel (Figure 5). This is a shallow, closed lake in a semiarid climatic zone; its area has fluctuated between about 1,950 and 26,000 km2 during the twentieth