Lakelevel Fluctuations

Throughout the arid and semiarid part of the world, there is commonly no runoff (surface water discharge) to the oceans. Instead, surface drainage may be essentially nonexistent (in areic regions) or it may terminate in interior land-locked basins where water loss is almost entirely due to evaporation (endoreic regions). In these basins of inland drainage (Fig. 7.18), changes in the hydrological balance as a result of climatic fluctuations may have dramatic effects on water storage. During times of positive water budgets, lakes may develop and expand over large areas, only to recede and dry up during times of negative water balance. Studies of lake-level variations can thus provide important insights into paleoclimatic

1800 f 16001400-

FIGURE 7.17 Fluctuations of the lower Grindelwald Glacier between 1590 and 1970 relative to the 1970 terminal position.The major recession since I860 (with minor interruptions around 1880 and 1920) is clearly seen. Moraines are indicated by bold curves (Messerli et al., 1978).

1200

g 1000

Q 600

200- /-glacier front extent of 1970

200- /-glacier front extent of 1970

FIGURE 7.17 Fluctuations of the lower Grindelwald Glacier between 1590 and 1970 relative to the 1970 terminal position.The major recession since I860 (with minor interruptions around 1880 and 1920) is clearly seen. Moraines are indicated by bold curves (Messerli et al., 1978).

conditions, particularly in arid and semiarid areas. In modern lake basins, periods of positive water balance are generally identified by abandoned wave-cut shorelines and beach deposits (Fig. 7.19) or perched deltas from tributary rivers and streams, and exposed lacustrine sediments at elevations above the present lake shoreline (Morrison, 1965; Butzer et al., 1972; Bowler, 1976). Periods of negative water balance (relative to today) are identifiable in lake sediment cores or by pa-leosols developed on exposed lake sediments (Street-Perrott and Harrison, 1985a). A study of the stratigraphy, geochemistry, and microfossil content of lake sediments from closed basins may be particularly valuable in deciphering lake history (Bradbury et al., 1981).

Most of the early studies of lake-level fluctuations focused on closed lakes where changes in the precipitation-evaporation balance led to volumetric changes and consequent adjustments in lake-level elevation. Recent work has attempted to expand beyond predominantly arid and semiarid regions, where such lake systems are commonly found, to higher latitudes where open (overflowing) lake systems are more common. In open system lakes, water balance changes are naturally compensated for by changes in outflow, but lake levels may also change somewhat, albeit far less than in closed lake systems. Where multiple sediment cores are available from a lake, it may be possible to detect in the sedimentary facies former shallow water conditions and to thereby date periods with lower lake levels (Harrison and Digerfeldt, 1993). Macrofossil and palynological evidence may also provide evidence of changes in lake level via shifts in the relative abundance of shallow and deepwater aquatic flora. Where such detailed studies have been carried out it may be possible to reconstruct water balance changes even in open system lakes (some of which may, of course, have become closed systems in drier periods). Although the vast majority of open system lake sediment studies do not

FIGURE 7.18 Areas of endoreic and areic drainage. Areic regions have no permanent surface drainage; endoreic areas are basins of inland drainage (Cooke and Warren, 1977; de Martonne andAufrere, 1928).

have a suite of cores from deep to shallow water, detailed study of sedimentologi-cal, macrofossil, diatom, and palynological data has enabled an interpretation of lake level changes to be made at some locations (Yu and Harrison, 1995; Tarusov et al., 1996). Lake-level changes derived in this way from open system lakes have then been used for large-scale hydrological reconstructions by Harrison (1989) and Harrison et al. (1996).

Lake-level fluctuations have been studied in dozens of closed basins throughout the world (for a list of principal works, see Street-Perrott et al., 1983; lake-level data are also available from the World Data Center for Paleoclimatology: see Appendix B). The majority of these studies are stratigraphie and provide only qualitative estimates of climatic conditions. Periods of higher lake levels are commonly described as pluvials, but the question of whether such conditions result from increased precipitation or lower temperatures and more effective precipitation (via reduced évapotranspiration) is controversial (Brakenridge, 1978; Wells, 1979). In an attempt to resolve the controversy, a number of studies have attempted to use the geomorphological evidence, together with empirically derived equations relating climatic parameters today, to make quantitative estimates of paleoclimatic conditions associated with particular lake stages in the past. These can be considered in two general categories: hydrological balance models; and hydrological-energy balance models.

FIGURE 7.19 Late glacial lake-shore terraces, Lake Tauca, Salar Ujuni, southwestern Bolivia.The highest shorelines (~70 m above the present-day lake level, see arrows) were formed ~ 13,000 yr B.R (photograph kindly provided by C.Ammann, University of Massachusetts).

7.6.1 Hydrological Balance Models

In a closed basin, variations in lake level are a function of water volume, which is in turn a reflection of the balance of water supply and water loss:

dV _ d(P + R + U) d(E + O) dt dt dt where V is water volume in the lake; P is precipitation over the lake; R is runoff from the tributary basin, into the lake; U is underground (subsurface) inflow to the lake; E is evaporation from the lake; and O is subsurface outflow from the lake. For any particular lake stage, if the hydrological balance is considered to be at equilibrium, such that then, P + R+ U = E + 0. Generally, subsurface inflow and outflow are considered to be negligible and are omitted from the equation, although in some cases they may be substantial; in the case of the Great Salt Lake, Utah, for example, subsurface inflow has been variously estimated at 3-15% of total lake input (Arnow, 1980). In most cases, however, the subsurface components are unknown even for modern lake levels and trying to estimate them for paleolakes would be extremely speculative. Assigning values of zero to these components, the hydrological balance equation for a particular basin is thus:

where Ah is the lake area; AT is the area of the tributary basin from which water drains to the lake; PT is mean precipitation per unit area over the tributary basin; k is a coefficient of runoff (hence PTk equals the runoff per unit area RT from the tributary basin); pl is mean precipitation per unit area, over the lake; and el is mean evaporation per unit area, from the lake. As only al and at are known for any given lake stage, the equation may be rearranged thus:

A solution of the equation therefore requires a knowledge of precipitation over the lake and adjacent catchment basin, runoff from the surrounding basin, and the amount of water that evaporates from the lake. All of these parameters are a function of many other variables that are also unknown, making a unique solution to the equation extremely difficult to say the least. To illustrate the uncertainties involved, Table 7.2 lists some of the principal factors affecting evaporation and runoff. Of major importance to both parameters is temperature, and this has enabled some limits to be placed on estimates of former runoff and evaporation values when reasonably good paleotemperature estimates are available. If paleotemperatures are known, empirically derived equations relating runoff, evaporation, and temperature (Figs. 7.20 and 7.21) can be used to solve the hydrological balance equation. However, these empirical relationships are often limited in the range of values considered, and constrained by inadequate or even nonexistent data. Consider, for example, the relationship between lake evaporation and temperature. Most empirical relationships are based on standard measurements of evaporation from metal pans 1.2 m in diameter, at different temperatures; lake evaporation is assumed to be less than pan evaporation by a factor of 0.7, based on empirical studies by Kohler et al. (1966). However, evaporation rates depend on a number of factors that have not been constant through time (Table 7.2), for example, lake volume and salinity variations. In large lakes, such as Lakes Superior and Ontario, there is a very poor correlation between monthly temperature and evaporation because much energy is used in raising the water temperature at depth (i.e., in heat storage).

Evaporation is at a maximum in fall and winter months when the lake surface eventually becomes warmer than the overlying air (Morton, 1967). Such an effect

TABLE 7.2 Factors Affecting Rates of Evaporation and Runoff

Evaporation

Runoff

Temperature (daily means and seasonal range)

Ground temperature

Cloudiness and solar radiation receipts

Vegetation cover and type

Wind speed

Soil type (infiltration capacity)

Humidity (vapor pressure gradient)

Precipitation frequency and seasonal distribution

Depth of water in lake and basin morphology

Precipitation intensity (event magnitude and

(water volume)

duration)

Duration of ice cover

Precipitation type (rain, snow, etc.)

Salinity of lake water

Slope gradients; stream size and number

0 0

Post a comment