190 Imprint of Climatic Zonation on Marine Sediments
Here, Test is a temperature estimate based on an assemblage of foraminifera (or other
shells), Pi is the proportion of the i-th species, and ti its temperature optimum. The
optimum, in this context, is the temperature at which we find the highest proportion
of the species in the calibration set.
The calibration set consists of sea-floor surface samples and corresponding surface water temperatures. For example, we may find that on the sea floor species no.
3 has its highest proportion underneath surface waters with an annual average temperature of 20 °C. We call this its optimum, and designate it as t3. For the temperature
estimate of a given sample, the product P3 . t3 is one of the terms of the above
summation. The larger the P3, the closer the sum will be to 20 °C. The presence of
species with a higher optimum temperature will pull the result up above 20 D C, and
vice versa. Under favorable conditions, this simple equation will estimate the correct
temperature to within better than 3 dc. The technique can be improved by considering the temperature range of each species, and by regression of the estimated temperature on the expected temperature, in the calibration set. The method then gives
excellent results. More powerful transfer techniques also exist, based on the same
principle of calibration. Factor-regression analysis is one of these techniques. It was
used by the CLIMAP group to map conditions in the ice age ocean.
7.2.3 Application of lransfer Methods: Climatic Transgression. Great progress
has been made in determining the temperatures of the sea surface for the latest
Pleistocene, especially in the North Atlantic. The information is retrieved from
numerous sediment cores distributed throughout the region. An example of such
reconstruction, by the CLIMAP group, is given in Fig. 7.4.
The distribution of surface water temperatures - and hence of surface currents and
water masses - differs greatly for the last ice age maximum (- 20000 years ago)
from the present situation.
GO°
SOO
30'
20°
10'
Cf
Fig. 7.4 a, b. Expansion of polar climate in the North Atlantic. 17000 years ago. a Today's se a
surface temperature in winte r. b Tran sfer map, winter surface temperature 17000 years ago (in
ocean) , and distribution of ice sheets and pack ice. [Data from CLIMAP in E. Seibold, 1975.
Naturwissenschafte n 62: 321 ; see also Science 191: 1131. 1976; Geol. Soc. Am . Mem . 145: 464 pp.
1976]
Here, Test is a temperature estimate based on an assemblage of foraminifera (or other
shells), Pi is the proportion of the i-th species, and ti its temperature optimum. The
optimum, in this context, is the temperature at which we find the highest proportion
of the species in the calibration set.
The calibration set consists of sea-floor surface samples and corresponding surface water temperatures. For example, we may find that on the sea floor species no.
3 has its highest proportion underneath surface waters with an annual average temperature of 20 °C. We call this its optimum, and designate it as t3. For the temperature
estimate of a given sample, the product P3 . t3 is one of the terms of the above
summation. The larger the P3, the closer the sum will be to 20 °C. The presence of
species with a higher optimum temperature will pull the result up above 20 D C, and
vice versa. Under favorable conditions, this simple equation will estimate the correct
temperature to within better than 3 dc. The technique can be improved by considering the temperature range of each species, and by regression of the estimated temperature on the expected temperature, in the calibration set. The method then gives
excellent results. More powerful transfer techniques also exist, based on the same
principle of calibration. Factor-regression analysis is one of these techniques. It was
used by the CLIMAP group to map conditions in the ice age ocean.
7.2.3 Application of lransfer Methods: Climatic Transgression. Great progress
has been made in determining the temperatures of the sea surface for the latest
Pleistocene, especially in the North Atlantic. The information is retrieved from
numerous sediment cores distributed throughout the region. An example of such
reconstruction, by the CLIMAP group, is given in Fig. 7.4.
The distribution of surface water temperatures - and hence of surface currents and
water masses - differs greatly for the last ice age maximum (- 20000 years ago)
from the present situation.
GO°
SOO
30'
20°
10'
Cf
Fig. 7.4 a, b. Expansion of polar climate in the North Atlantic. 17000 years ago. a Today's se a
surface temperature in winte r. b Tran sfer map, winter surface temperature 17000 years ago (in
ocean) , and distribution of ice sheets and pack ice. [Data from CLIMAP in E. Seibold, 1975.
Naturwissenschafte n 62: 321 ; see also Science 191: 1131. 1976; Geol. Soc. Am . Mem . 145: 464 pp.
1976]