Gepl.narod.ru

Transactions (Doklady) of the Russian Academy of Sciences/Earth Science Sections, Vol. 348, No. 4, 1996, pp. 626–629.
Translated from Doklady Akademii Nauk, Vol. 348, No. 1, 1996, pp. 111–114.
Original Russian Text Copyright 1996 by Klimenko, Klimanov, Fedorov.
English Translation Copyright 1996 by åÄàä ç‡Û͇ /Interperiodica Publishing (Russia). GEOPHYSICS
The History of the Mean Temperature of the Northern
Hemisphere over the Last 11000 Years
V. V. Klimenko*, V. A. Klimanov**, and M. V. Fedorov*
Presented by Academician G.S. Golitsyn January 24, 1995 The magnitude and the sign of future climate all large-scale (about one degree) climate variations in changes are determined by a complex superposition of Late Glaciation–Holocene are manifested in all parts of anthropogenic and natural factors. There are a number the globe with a remarkable synchronism but different of reasons to believe that over the last centuries the cli- amplitude. Based on this conclusion and on the abun- mate has been cooling and will continue to do so in the dant temperature variation data on vast land regions, we future [1–4]. This can, to a some extend, reduce the rel- attempted to reconstruct global climate variations over ative temperature rise caused by anthropogenic activity.
Hence, the quantitative study of regularities of the nat-ural evolution of climate has a great significance.
The principal obstacles in obtaining the time series Attempts to estimate the parameters of natural climate of mean global temperature from palynological data are variations were repeatedly made in the past [5, 6].
the following: initial series for specific palinological However, the lack of detailed information on the mean sections differ in duration as well as in time step; the annual global and hemispheric temperatures for a suffi- sampling (section) sites are chaotically spaced.
ciently long timespan was, up to the present, a seriousobstacle that prevented one from making generaliza- To overcome the first obstacle, we reduced all series tions. In this work we attempt to reconstruct the tem- to a standard time scale (at 25-year intervals) and a sin- perature series for the Northern Hemisphere during the gle period (11 ka B.P.—1950). When performing this Late Glaciation-Holocene from paleoclimatic data.
procedure, we took into account results of the compar- Results of modern instrumental observations were also ison of temperature series obtained at different sec- tions, which demonstrate substantial compatibility For the paleoclimatic reconstruction of Late Glacia- despite the large distances between them. The fact that tion and Holocene, we used the informational-statisti- the studied series contain information on rather long- cal method [7] based on the statistic relation between term and large-scale temperature variations only is, in the contemporary spore and pollen spectra and climatic our opinion, the reason why these series demonstrate conditions (mean temperatures for July and January, consistent temperature changes. But the scale of these mean annual temperatures, and total annual precipita- variations is different for different sections. As a rule, tion). The accuracy of determinations of mean values isas follows: July and mean annual temperatures, their amplitude increases with an increase in the geo- ±0.6°C; January temperature, ±1°C; and total annual graphical latitude of the section; this is in good agree-precipitation ±25 mm.
ment with conclusions obtained from the instrumentalclimate observations in [9]. Hence, we assume that More than 60 paleoclimatic curves with different degrees of detail were drawn out by this method for each of the reconstructed temperature series T ' can be plane regions of northern Eurasia. Figure 1 illustrates the reconstructed mean July temperatures for someregions located in temperate and subpolar latitudes and T ' = T (t) + ε (t), spaced many thousands of km apart. The rather goodcorrelation between all these curves is an additional where Ti(t) is the mean annual air temperature over the argument in favor of the recent concept [2, 8, etc.] that section i at instant of time t and εi(t) are the accuracy ofmeasurements. Note that for any two sections i, j, func-tions T *Institute of Nuclear Safety Problems, Russian Academy i(t) and Tj(t) are connected by a linear relation.
of Sciences, B. Tul’skaya ul. 52, Moscow, 113191 Russia In order to determine the shape of the signal T(t) asexactly as possible, we selected the nine most detailed **Institute of Geography, Russian Academy of Sciences, Staromonentnyi per.29, Moscow, 109017 Russia series, the duration of which exceeds the chosen time THE HISTORY OF THE MEAN TEMPERATURE OF THE NORTHERN HEMISPHERE Fig. 1. Paleoclimatic curves (the deviations of the mean June temperatures from the present–day ones (in °C) for different regions
of northern Eurasia: (1) Karelia, (2) central Western Siberia, (3) central Yakutia, (4) central Byelorussia, (5) Bashkiria, (6) Primorsk
region.
interval, and constructed the averaged temperature bal climate is followed by a very nonuniform spatial distribution of climatic parameters, the mean tempera-tures for rather large regions can be calculated with rea- sonable accuracy by means of a linear transformation of the mean global temperature. In other words, the fol- If we assume, additionally, that Coν(ε , ε signal to noise ratio for this series must be substantiallygreater than that for the initial ones. In a first approxi- r(t), Tgl(t), and (t) are the mean regional tem- mation, we can neglect the second term in (2) and con- perature in the year t, mean global temperature in the sider that the averaged series does not involve a noise.
year t, and random noise, respectively. It is evident thatthe greater the size of a region, the more precise will be We estimated coefficients of the linear regression of the model (3). Therefore, the coefficients of correlation each initial series on averaged series (2). The closing of between the mean global temperatures and tempera- gaps and the extrapolation of initial series was carried tures for both the Northern and Southern Hemispheres out by substitutions of corresponding points of aver- calculated from the data of Jones et al. [11] are equal to aged series (2) transformed with consideration taken 0.95. But model (3) may be of use for considerably smaller regions, as was demonstrated in [9] for the 10° In order to overcome the second obstacle, the results latitude by 360° longitude zone of the Northern Hemi- of temperature measurements at chaotic sampling sphere (data on instrumental observation of climate).
(section) sites were interpolated into zones of a regular In our work, the β coefficient was obtained by aver- 5° latitude by 10° longitude grid. We chose the interpo- aging (with weights proportional to area) correspond- lation technique of Jones et al. [10] for compiling the ing coefficients for a 10° latitude by 40° longitude global grid archives of air temperatures. In accordance region (in our works, for regions 10° lat. by 40° long. in with this technique, the temperature in a grid zone is size) of the former USSR. In this case, the value of β is determined by the sum of temperatures measured at equal to 1.47. Therefore, we used the following formula nearest points with weights that are inversely propor- for calculating the global average temperature: tional to their distance from the zone. The grid dataobtained in this way were integrated for the territory of the former USSR. In our opinion, the series obtained constitute important information on the change in theglobal climate and can be used, in particular, for recon- where β = 1.47, and the constant α was chosen in such structing the time series of the global mean temperature a way that the mean value of the series obtained over of surface air. Actually, although the change in the glo- the 1851–1950 period coincided with the mean hemi- TRANSACTIONS (DOKLADY) OF THE RUSSIAN ACADEMY OF SCIENCES/EARTH SCIENCE SECTIONS Vol. 348 No. 4 1996 Holocene warming differed from present-day ones by only few decigrades. This, however, seems natural if we take into account the preservation of much of the conti- nental ice and cold North Atlantic in this period.
The temperature curve obtained is valuable initial material for constructing a long-range prediction of thebehavior of natural climate. In doing so, our first objec- tive was to determine the temperature effects of longperiod (about 104–105 years) cycles of the natural cli- mate, which, apparently, are related to alterations of the parameters of Earth’s heliocentric orbit. In order to determine the temporal trend of the cli- mate change, we used the standard procedure for timeseries analysis [15]. The study of the history of several interglaciations let us to choose a quadratic function forestimating the temperature change trend. The coeffi- cient of correlation between the relation Fig. 2. Change of mean temperature in the Northern Hemi-
sphere (expressed as anomalies relative to 1951–1980) over (t) = At + Bt + C, and experimental data reaches 0.93; Y(t) is the trend estimate of a temperature change; t is time; A, B, and C spheric temperature anomaly over the same period (−0.25°C) derived from instrumental observation [11] Our results show that the rate of mean global tem- (previously, Jones’s series was recalculated relative to data perature drop caused by orbital factors will be of the for 1951–1980). Thus, the obtained T order of (3–6) ⋅ 10–4 °C per year during, at least, the reconstruction of anomalies (relative to 1951–1980) in the next few thousand of years. This value is an order of mean annual temperature of the surface air in the magnitude less than the rate of the temperature rise reg- istered by instrumental observations during the lastcentury and assumed for the 21th century. Neverthe- This reconstructed series depicts a complex pattern less, the trend, which is well pronounced and stable in of alternating cooling and warming periods during Late amplitude and sign, suggests that the significant tem- Glaciation–Holocene. We cannot agree with the state- perature effect resulting from orbital variations should ment [12] about a remarkable climatic stability over be taken into consideration for correct forecast and the the last 10 ka, which is based on recent results from the long-range (over several centuries) climatic reconstruc- investigation of the Greenland ice core at the Summit tion. The aforesaid statement refers equally to the cal- station. The onset of temperature extremums are in culation of anthropogenic warming, the significance of good agreement with the existing notions based on iso- which will be progressively diminished due to the lated local reconstructions. One should not consider this fact surprising because of the suggestion men-tioned above about the global character of large-scaleclimate variations. As for temperature variation ampli- tudes, the pattern presented in Fig. 2 is, in contrast, This work was supported in part by the Mining inconsistent with conventional notions [13]. Perhaps Industrial Corporation (Russia) and U. S. Department only the big Atlantic optimum (6–5 ka B.P.) retains its of Energy, project no.1753-300224. V.V. Klimanov significance as the warmest and, simultaneously, the acknowledges the Alexander von Humboldt Founda- most continuous Holocene period. Mean temperatures tion (Germany) for invariable support and attention.
and the peak values exceeded present-day values by0.82°C and 1.4°C. Our estimate of the mean tempera-ture is in good agreement with calculations (0.6–0.7°C) in [14] conducted on the basis of paleoclimtic recon- 1. Gerasimov, I.P., Meteorol. Gidrol., 1979, no. 7, pp. 37–53.
struction for a nontropical zone of the Northern Hemi- 2. Klimanov, V.A., in Paleoklimaty pozdnelednikov’ya i golotsena (Paleoclimates of Late Glaciation and Our estimate is perhaps more valid because it is Holocene), Moscow: Nauka, 1989, pp. 29–33.
based on a distinct pattern of latitudinal temperature 3. Klimenko, V.V., Klimenko, A.V., Snytin, S.Yu., and distribution anomalies registered by reliable instrumen- Fedorov, M.V., Teploenergetika, 1994, no.1, pp. 5–11.
tal observations over almost 150 years. The subboreal 4. Berger, A., Vistas Astron., 1980, vol. 24, pp. 103–122.
(4.2–3.3 ka B.P.) maximum was the second highest 5. Johnsen, S.J., Dansgaard, W., Clausen, H.B., and Lang- temperature value; temperatures during the Early way, C.C., Nature,1970, vol. 227, no. 5257, pp. 482–483.
TRANSACTIONS (DOKLADY) OF THE RUSSIAN ACADEMY OF SCIENCES/EARTH SCIENCE SECTIONS Vol. 348 No. 4 1996 THE HISTORY OF THE MEAN TEMPERATURE OF THE NORTHERN HEMISPHERE 6. Lamb, H.H., Climate, History, and the Modern World, 11. Trends 91: A Compendium of Data on Global Change, Eds. Boden, T.A., Sepanski, R.J., and Stoss, F.W., OakRidge: Carbon Dioxide Information and Analysis Cen- 7. Klimanov, V.A., Vestn. Mosk. Univ., Ser. 5. Geogr., 1976, 12. Dansgaard, W., Johnsen, S.J., Clausen, H.B. et al., 8. Borzenkova, I.I.,Izmenenie klimata v Kainozoe (Climate Nature, 1993, vol. 364, no. 6434, pp. 218–220.
Change in Cenozoic), St. Petersburg: Gidrometeoizdat, 13. Climate Change: The IPCC Scientific Assessment, Cam- bridge: Cambridge Univ. Press, 1990.
9. Vinnikov, K.Ya., Chuvstvitel’nost’ klimata (Climate 14. Velichko, A.A. and Klimanov, V.A.,Izv. Akad. Nauk Sensitivity), Leningrad: Gidrometeoizdat, 1986.
SSSR, Ser. Geogr.,1990, no. 5, pp. 38–52.
10. Jones, P.D., Raper, S.C.B., Bradley, R.S. et al. Climate and 15. Bendat, J.S. and Piersol, A.G., Random Data: Analysis Applied Meteorology, 1986, vol. 25, no. 2, pp. 161–179.
and Measurement Procedures, New York: Wiley, 1986.
TRANSACTIONS (DOKLADY) OF THE RUSSIAN ACADEMY OF SCIENCES/EARTH SCIENCE SECTIONS Vol. 348 No. 4 1996

Source: http://www.gepl.narod.ru/Articles/Klim96.pdf