Quantitative Analysis of Climate Change and Human Crises in History - Space-Time Integration in Geography and GIScience - page 241

Table 14.3 Regressions of various population checks in the Northern Hemisphere on time
and Northern Hemisphere temperature anomalies in AD 800-1900. Population checks and
temperature anomalies data have been smoothed by 100-year Butterworth low-pass filter prior to
statistical analysis. The regressions are corrected for autoregressive disturbances using the Prais-
Winsten estimation method. Elasticity can be interpreted as the percentage of population check
changes in response to a one percent increase in temperature. t calendar year divided by 10 3 . D

R 2 adj D adjusted R 2 calculated for the untransformed variables

Dependent Independent variable

variable Constant t t 2 t 3 Temp Elasticity R 2 adj

Famine 4.445 6.713*** 1.441*** 3.493*** 0.212 0.309

Famine 0.383 0.936** 3.555*** 0.216 0.279

Famine 0.205 0.386** 3.580*** 0.217 0.298

War 9.333 2.783 1.164 3.678*** 0.119 0.226

War 7.651 0.180 3.653*** 0.118 0.218

War 7.774 0.071 3.657*** 0.118 0.218

Epidemics 7.643 18.248*** 9.772*** 1.680*** 0.113 0.927

Epidemics 3.398 3.328*** 1.514*** 0.103 0.886

0.113 0.914
*Significant at 0.05 level (2-tailed) ( P < 0.05); **Significant at 0.01 level (2-tailed) ( P < 0.01);
***Significant at 0.001 level (2-tailed) ( P < 0.001)

Epidemics 2.120 1.559*** 1.670***

We also found a very significant correlation between the Northern Hemisphere
temperature swings and population growth rates for the world, Northern Hemi-
sphere, Asia, North America, and Europe, as well as for China for the period
AD 1400-1900 17 (Table 14.4 )(Zhangetal. 2007a ). The correlation coefficients
for Europe and North America are lower than the others. The reason for this is
that Europeans and Americans lived in an ecologically open system that began
during the Industrial Revolution, one in which European population pressure could
be relieved by cross-continental colonization and migration. When the nineteenth
century (the age of Great Migration in Europe) population data were not included
in the analysis, the correlations for Europe and North America were almost as high
as those for Asia and the whole Northern Hemisphere. Similarly, the large-scale
trans-Atlantic migration from Europe to North America in AD 1700-1900, which
is supposed to be driven by cooling, can also be substantiated by the significant
negative correlation between the Northern Hemisphere temperature anomaly and
North America's population growth (r D 0.341, P < 0.001).
In addition, we explored quantitatively the temporal pattern, spatial pattern, and
triggers of population collapses 18

in relation to climate change at the global scale

17 See footnote no. 6 for the details of Northern Hemisphere temperature change. The calculation
of population growth rate was based on McEvedy and Jones ( 1978 ). Prior to statistical analysis,
all data were smoothed by the 40-year Butterworth low-pass filter to remove fluctuations on time-
scales <40 years.
18 The identification of population collapses was based on McEvedy and Jones ( 1978 ).