A chronology prepared by Sindh historian M.H. Panwhar has some interesting climate-related entries for our period of interest:
'Famine in Gujarat and Deccan due to failure of monsoons. This famine was due to failure or rains in 1630 and excessive rain in 1631. People sold their children so that they may live. . . . Hides of catle [sic] and flesh of dogs were eaten, cremated bones of dead were sold with flour and cannibalism became common. . . . Three million people died between 1630-1633 in Gujarat.
'1636-1637. Punjab had famine. . . .
'1640. Heavy rain caused floods and destroyed crops in the Punjab and Kashmir, causing famine. . . .
'1640-44. Rains failed continuously in many parts of Northern India and famines occurred in Agra province. . . .
'1642. Famine occurred due to heavy rain and floods in the Punjab. . . .
'1646. Drought in Agra and Ahmedabad. . . .
'1647. Rains failed in Marwar. Famines, high mortality. . . .
'1648. Failure of rains in Agra area. . . .'
Newson (35) writes, 'there is some evidence, notably from tree-ring data from Java, that during the seventeenth century some countries in Southeast Asia experienced unstable climatic conditions, perhaps linked to the 'Little Ice Age' in Europe, that included frequent dry periods that resulted in food shortages and famines. However, Peter Boomgaard suggests that climatic conditions were probably not so anomalous . . .'
China experienced severe cold in 1629-43, as well as severe drought in 1637-43 (Brooks 269). Not coincidentally, the Ming dynasty fell in 1644. Climate extremes led to famine (the first big one was in 1630), which led to peasant revolts.
Famine also meant greater vulnerability to disease, and migrations to flee affected areas aided disease transmission. There were significant epidemics in the northwest beginning in 1633, and in the Yangtze valley in 1639. (Brooks 250ff).
In the Yangtze Delta, historical records show that 1635-1644 was marred by four flood events (years?) and six drought events. This was part of a larger trend; there were many flood and drought events in the period 1540- 1670. In contrast, in 1495-1504, there was one flood event and no droughts (Jiang). Qiang says in 1550-1850, calamities 'occurred by turns and sometimes, both drought and flooding occurred in the same year.' There was snowfall and frost on low ground in south China in 1635 and 1636, and the River Huai froze over in 1640 (LambCPFF 612).
Perhaps the most unique aspect of the Little Ice Age in China was the increase in reports of dragon sightings (Brooks 6ff); for example, 'two dragons were spotted in autumn of 1643. . . .' (14). Dragons were associated with water, and thus with storms and, more generally, bad weather. But they were both symbols of the emperor and celestial messengers. If people were seeing dragons, then what they were really perceiving was climatic evidence that the emperor had lost the Mandate of Heaven. And of course these reports in turn made it more likely that the emperor would lose support.
In Japan, 17th-century (and earlier) temperatures have been reconstructed on the basis of
– the dates of cherry blossom viewing parties in Kyoto
Overall, the average full-flowering date was day 105; the average for the 17th century was 106 (April 16), and for the 20-21c, 101 (April 11). The estimated March mean temperature for the 1630s was around 7oC, and the LIA low was around 6oC in late-17th century and early 18th century. A deeper low of ~5oC was inferred for the early-14th century)(Aono). In 1633, the cherry trees blossomed on April 8 (LambCPFF 607).
– the dates of freezing (December-January), buckling (
Overall, the average freezing date was January 15, and the mean for the 1630s was about 12 days early, and that was actually one of the coldest decades recorded)(Lamb 256). By way of example, the dates were 1634: Jan. 9, 1635: Dec. 28, 1636: Jan. 2, 1637: Jan. 11, 1638: Dec. 31, 1639: Jan. 21 (LambCPFF 609).
There is a correlation between the mean winter temperature at Tokyo (Edo) and the Lake Suwa freezing date; the estimated temperature is 4.1oC for the 1630s and 1640s (LambCPPF 610).
– the first snow cover in Tokyo
This was January 6 (1633), December 16 (1638), February 2 (1640), November 28 (1642), and January 10 (1648). (611).
– the proportion of a cold-adapted species in pine pollen from Ozegahara, a raised bog 150 km north of Tokyo
Most of the 17th century appears to have been a bit on the warm side (Batten 18).
– tree ring data
That from Yaku Island in the south shows two sharp temperature drops in the 17th century, and from central Honshu shows slow decline in temperature during 17th century)(Batten 19, 21).
Generally speaking, the winters in central Japan were most severe in 1500-1520, 1700-10, and 1850-80, not in the period of interest to us now (LambCHMW 227).
While the Genroku (1695-6), Tenmei (1782-7) and Tempo (1833-39) famines all occurred during particularly cold periods, Japan's population still doubled from 1600 to 1721 (Batten, 57, 59).
The Spanish take advantage of wind (midlatitude westerlies, subtropical northeast trades) and currents (the North Pacific gyre) in the Manila-Acapulco galleon trade. Spanish archives show that the average duration of the Acapulco-Manila passage (westing made mostly around 12oN latitude) increased steadily from 80 days in 1600 to 100 in 1640 and a peak of a little over 120 days in 1655, then descended to gradually to a plateau of 90-100 days in 1690-1750. 'Virtual voyage' calculations indicated that the slowing was most likely the result of a northeastward shift in the position of the 'southwest monsoon trough' (the ICTZ) in June (Garcia-Herrera).
The fictional cosmic event we call the Ring of Fire replaced a six mile diameter of 1631 Thuringian air with one from 2000 West Virginia. What we will speculate about in this part is just how profound an effect this event would have had on weather (short-term) and climate (long-term), both locally and remotely.
Okay, folks, hold your hats. It's time for our intellectual roller coaster to plunge into the abyss of chaos theory. Just be thankful that I am sparing you the mathematics that I studied, and that I am concentrating on the implications.
Let's begin the descent gently by talking about the origin of the term 'butterfly effect.' It comes from the title of a presentation by the mathematical meteorologist Edward Lorenz: 'Does the flap of a butterfly's wings in Brazil set off a tornado in Texas?'
This was a reference to his observation of chaotic behavior in his atmospheric convection model. Chaotic behavior means that a small perturbation in the initial conditions results, eventually, in a large and to some degree unpredictable divergence ('bifurcation') in the 'final' state. This chaos was not the result of randomness; Lorenz' model was completely deterministic. However, the chaos had the
Chaotic behavior can be inherent in the actual physical system-and the existence of nonlinear relationships is necessary for chaotic behavior to emerge. It can also be an artifact of numeric evaluation of a mathematical model of a physical system. (The input data is for grid points, not spatially continuous; the evolution of the model is calculated in discrete time steps, not continuously, which means that we are working with finite differences not true derivatives; and there will be rounding errors inherent in how computers handle numbers.) Of course, it isn't necessarily easy to separate the two!
The nonlinear dynamics of the climate system include both positive and negative feedback loops. As an example of positive feedback, increasing the surface temperature of land or water just below the freezing point results in conversion of snow and ice (high albedo) to bare earth or liquid water (low albedo), which increases
