Jökull - 01.01.2005, Blaðsíða 12

Wallace S. Broecker
by catastrophic additions of fresh water to the north-
ern Atlantic. One such mode of addition was through
the melting of huge armadas of icebergs launched into
the northern Atlantic as the result of a collapse of
the ice dome centered over Hudson Bay (Hemming,
2004). Another was gigantic floods associated with
the sudden escape of fresh water impounded in lakes
in front of the retreating Canadian ice sheet (Broecker
et al., 1989). Today, only the Greenland ice cap holds
an amount of “poised” fresh water capable of trigger-
ing a sudden conveyor shutdown. However, none of
the models of Greenland’s response to global warm-
ing suggest a sudden massive collapse into the sea.
Rather, just as the amount of high latitude precipita-
tion will slowly increase, so also will the rate of melt-
ing of Greenland’s ice.
WATER RELATED CRISIS
If the above assessment is correct, can we assume
that no large bumps lie along the greenhouse warm-
ing road? Perhaps, but as for the last half-million
years the Earth has not experienced climates signifi-
cantly warmer than today’s, we are about to enter un-
known territory. If surprises are in store for us, my
guess would be that they will involve redistribution
of rainfall (i.e., droughts and floods). Climate mod-
els suggest that as the world warms, the tropics will
receive ever more rainfall and the world’s desert re-
gions will receive ever less. While we lack warm ana-
logues, we do have a well-documented cold analogue.
During glacial time closed basin lakes in the deserts
of western North America, the Middle East, north-
western China and Australia were several times larger
than now. While evidence from the tropics is sparse,
it has been shown that Lake Victoria was dry during
late glacial time (Johnson et al., 1996) and that the
other lakes in the Africa’s rift valleys were consider-
ably smaller than they are today. Thus, it appears that
when the Earth was colder the tropics got less rain-
fall and the desert regions got more. Thus, the paleo
record adds credence to the prediction by models.
But one might ask whether these changes in mois-
ture delivery will be gradual or is there a chance
that they will come upon us suddenly? Although no
firm answer can be given to this question, regional
droughts during the present interglacial offer some
important clues.
The most dramatic of these droughts are two
which hit the dry lands of western North America dur-
ing the Medieval Warm interval, (i.e., the time when
Eric the Red and his band of Vikings colonizedGreen-
land). These western North America droughts are dra-
matically recorded by the trunks of dead trees which
project above the water level in present day lakes,
rivers and swamps in the mountains along the west-
ern border of the desert (Stine, 1994). Counts of an-
nual rings reveal that some of these trees survived for
150 years. Radiocarbon dates place them in two time
intervals, 1050 to 1200 and 1300 to 1375 AD. Of par-
ticular interest are the tree trunks projecting through
the surface of Lake Tenaya in the high country near
Yosemite Valley. These trees are rooted 13 meters be-
low the bed rock sill which serves as the lake’s outlet.
Over the last 130 years only once did this lake fail to
overflow in response to the input of snowmelt. Yet for
two century-long intervals during the time of the Me-
dieval Warm, it was never once filled to capacity. In-
stead, the reduced inflow must have been lost entirely
by evaporation. We can state this with confidence for,
had the lake risen to its sill level and overflowed, it
would surely have killed the trees.
Fossil tree trunks from a number of other locali-
ties tell similar stories (for one of these see Figure 9).
Clearly, western dry lands experienced two century-
duration droughts more intense than any of the three
to eight-year-duration droughts which have occurred
during the last 100 years. This comes as a surprise
for, as far as we can tell, climate during the Medieval
Warm was not significantly warmer than today’s.
But, this leaves open the question as to whether
these droughts began and ended suddenly. At present,
it can only be said that the two medieval droughts
were separated by a time when closed-basin Owens
Lake and Mono Lake which occupy the desert just
east of the steep front of California’s Sierra Nevada
mountain range both achieved high stands (Stine,
1990). Research is currently underway to determine
what dendro-correlated tree-ring thicknesses have to
say about the onset and demises of these mega-
droughts.
12 JÖKULL No. 55, 2005