Climate Reconstructions: Post-RoF Grantville Seasonal Average Temperatures
The place of greatest interest to the up-timers is, of course, the location in Thuringia where the RoF deposited Grantville. The center of the RoF was at approximately 11o16' east longitude, 50o40'12" north latitude. The closest Luterbacher grid point is 50.75N, 11.25E, and the reconstructed seasonal temperatures for this location are in Table 2-5A (with comparison to pre-RoF Grantville at the bottom of the table).
It can be seen that in OTL 1620-49, the growing season (April-September) was probably about 4-5oC. colder than in 1971-2000 Grantville. If extremes moved downward the same amount, that probably wouldn't shift Grantville into a new plant hardiness zone (that would require an 18oC change.)
Climate Reconstructions: Magdeburg Seasonal Average Temperatures, 1630-39
Magdeburg is at 52o07'N, 11o38'E, and the closest grid point is 52.25N, 11.25E. The data for that grid point are in Table 2-5B.
It is possible to extract the reconstructed temperatures for other locations in Europe, too, given their latitude and longitude. The necessary data set and format information are here:
http://www.cru.uea.ac.uk/cru/projects/soap/data/recon/#luter04 Please note I had to write a program to extract the data, because Excel can't import 18,000 columns of data. . . .
Climate Reconstruction: Plausible Grantville Monthly Average Temperatures
Unfortunately, I don't have gridded monthly temperature reconstructions covering the 1630s. Mark Twain once said, "there are three kinds of liars: liars, damn liars, and statisticians." We can make an educated guess as to what the monthly temperatures were, using statistics for other time periods. There are a number of ways that this can be done. I assumed that the relationship of monthly to seasonal temperatures for Grantville was the same as for Central Europe.
Or, in mathematical terms,
Grantville average for that month= Grantville average for that season (from LuterbacherTemp) + adjustment, where the adjustment was Dobrovolny's central Europe average for that month – central Europe average for that season.
Table 2-6A provides my reconstructed monthly temperatures for the location that Grantville was transported to. For convenience, I also repeat the climatological normals for Fairmont, West Virginia.
Bear in mind that these monthly numbers, even if accurately reconstructed, are the likeliest climate statistics to be corrupted by the "butterfly effect" of the RoF. So that's another good reason to view them as general indications rather than gospel truth.
What I thought most noteworthy about them was how fast temperatures dropped off during autumn and rose during spring.
Table 2-6B provides the average and standard deviation, over 1766-1850, of Luterbacher's reconstruction of monthly mean temperature at the same location. These, of course, reflect a different time period, but they save us the trouble of trying to convert temperature anomalies into absolute temperatures.
Grantville Growing Degree Days
We can estimate the number of growing degree days for a given location and year, for whatever base is appropriate to a particular crop. (Schenkler; Thom). Using the estimated monthly means and standard deviations from Table 2-6A, then by Thom's method we get the dramatic results shown in Table 2-7.
Grantville Extreme Temperatures
We also can make some educated guesses as to typical daily minimum (usually nighttime) and maximum (usually daytime) temperatures. The climatological norm of the monthly mean of the daily temperature range (maximum-minimum, DTR) varies from month to month, and is affected by latitude (which determines solar radiation variation), distance from the shore (affecting exposure to sea breezes and degree of low-level cloudiness), and precipitation. In the northern hemisphere, the DTR peaks at 20-40°N. In modern Europe, at latitude 50°N, the DTR is about 12°C in July and 7°C in January (Geerts Fig. 3). Moving inland more than 100 km increases the DTR by about 2°C. Those rules should apply to Grantville in Thuringia.
Unfortunately, there's a catch: the DTR can change as the climate changes. In particular, "the blocking action of greenhouse gases would be most effective where outward radiation was most important for cooling the Earth: warming would come especially at night" (Weart). So the DTR of the late-twentieth century is probably smaller than that of the seventeenth, when greenhouse gases were at lower concentrations and thus had less of an upward influence on nighttime temperatures.
****
The USDA plant hardiness zones are defined on the basis of the climatological norm of the annual minimum-the lowest daily minimum recorded during the course of the year. Unfortunately, to calculate that, by Monto Carlo methods, we need to know not only the monthly mean and standard deviation for (at least) January, but also the correlation of one day with the next.
Plant Hardiness Zone Maps have been created for modern Europe that use the same scale as the USDA maps (Heinze), and based on 1881-1930 data (although Lorek says that including 1931-1960 data would have an insignificant effect); Grantville post-RoF is in zone 6b (average January minimum of 20.5-17.8oC, -5 to -10oF) and Magdeburg in 7b (14.9-12.3oC, 10 to 5oF).
****
A change in climate may simply shift the mean temperature, and leave the variability intact. Or it can alter the variability, too. If for example, it was not only colder in the 1630s than in the 1990s, the variability increased, then the likelihood of frosts would be much greater than you would expect by just considering the mean. To complicate matters further, there's no guarantee that the "cold" and "warm" tails of the distribution will be affected by the same amount or even in the same direction.
Scientists have looked at both twentieth-century observational data (Kurbis, Moberg, Easterling, Karl, Michaels) and runs of global climate models, and about all I can conclude from this is that it's not safe to assume that the variability of temperature was the same in the early-17th century as it was in the modern period. Unfortunately, I have no easy way to predict how it was different and therefore I must just admit that this is something the authors may easily play around with.
Climate Reconstructions: European Seasonal Precipitation, 1630-39
Temperature, of course, is only one of the two principal dimensions of climate; the other is precipitation. Pauling reconstructed precipitation across Europe in 1500-1900, and his 1630-39 results are in Table 2-8. Fig. 3 compares the average for 1630-39 to that for 1990-99.
I provided the standard deviation, as well as the average, for 1630-39, and it is apparent that the annual variation from year to year for a given season was fairly small. However, because of the North Atlantic Oscillation, it's not unusual for northern Europe to be dry when southern Europe is wet, and vice versa.
Climate Reconstructions: European Seasonal Sea Level Pressure Patterns, 1630-39
Reconstructions are available of the average sea level and upper air pressure (in millibars) for each season in 1500-1658, and thereafter monthly (Luterbacher2002). For reasons I will explore in part III, this pressure history is more likely to get buffeted by the RoF than the temperature data, so don't put a lot of faith in it. Still, the patterns shown will remain plausible patterns. For 1630-4 and 1636-39, we had a winter high near the Azores and low near Iceland, with these highs and lows weaker in the other seasons. The Icelandic Low was distinctly weaker and displaced in the winter of 1635.
Radio Communications in the 1630s
Sunspot number normally varies according to a somewhat irregular 11-year cycle. However, there have been several periods of prolonged depression of sunspot number, notably the Sporer Minimum (1460-1550), the Maunder Minimum (1645-1715), and the Dalton Minimum (1790-1830), all of which correlate with cold temperatures. In the old time line, of course, no one had to worry about radio communications during these minima!
While it may sound as though we don't have to worry about the Maunder Minimum yet, that's not so. You see, the sunspot number is already on the decline. The total number of sunspots in the period 1630-1640 (eleven years) was 185.4. In
contrast, in 1980-1990, the total was 956.1. The period 1645-1715 was simply worse, with no sunspots at all in most years.
So the decade of the 1630s may be considered the "slippery slope" down to the Maunder Minimum. And that means that (with the exception of 1638-9 and 1642) we will be facing progressively greater limitations on the range of radio communications.
The Arctic
The great circle route is the shortest distance between two points, but it can require you to sail at dangerously high latitudes. While the cold can cause frostbite and sap strength, the greater danger is from sea ice.
In the North Atlantic, the principal pathways by which ice can descend from the Arctic Circle are the Davis Strait and Baffin Bay between Newfoundland and Baffin Island to the west and Greenland to the east, the Denmark Strait (Greenland Sound) between Greenland and Iceland, and the Norwegian Sea between Iceland and Norway. Nowadays, Arctic sea ice reaches its maximum overall extent in March (Polyak), and I suspect this was likewise true in the 17th century. However, note sea ice can expand in the Greenland sound while contracting in the Davis Strait, and vice versa. In a "normal" severe year, the ice could surround Spitsbergen and bisect Iceland. It only rarely reaches the south coast of Iceland. (Ogilvie).
Looking at the documentary evidence-based sea-ice index for Iceland, filtered through a 15-year low-pass filter, the 1630s and 1640s exhibited values under 2, while the filtered index climbed above 5 in the 1780s, 1810s, and 1830s. (Ogilvie). So the 1630s are not especially bad insofar as sea ice is concerned, although in one of the years (1632? 1633?) it jumped to level 6. An earlier (Koch 1945) study says that there was drift ice at the coast for 24 weeks in 1633, 26 in 1638 and 17 in 1639 (LambCPFF 583). Consistently, GroveLIAAM (22) states that "the 1630s . . . which were cold on land, saw little sea ice. . . ."
For the waters around Greenland, based on GISP2 ice core chemistry, the sea ice levels increased more or less steadily throughout the first half of the 17th century. However, the levels in the early- and mid-19th century were higher (Dugmore Fig. 7).
Jabe McDougal's musings continued, "After the high school had been saved from the Croat raiders, there had been a wave of interest in Swedish and Scandinavian history. Jabe had learned about the Little Ice Age and its presumed role in the death of the Viking colonies in Greenland." (This role is now considered debatable, see e.g. Mann.)
While we aren't interested in colonizing Greenland, it does have resources of interest. In year 1633 of the new time line, the Dutch metal and armament magnate Louis de Geer sent an expedition to Ivigtut (61o12'N/48o10'W) in southwest Greenland, at the mouth of Arsuk Fjord, to search for cryolite (the flux needed for efficient electrolytic production of aluminum metal from aluminum oxide) (Mackey, Kim, "Land of Ice and Sun," Grantville Gazette 11). The most obvious objection to an expedition of this type would have been that the local climate, oppressively cold even today, would have been far worse during the Little Ice Age.
Well, maybe. But as Kim pointed out when the story was in slush, while it was certainly cold, there was evidence that in the 1630s, it was no colder than when cryolite mining began (1854). Production was 14,000 tons in 1857-67; 70,000 in 1867-77 (Johnson's).
While I don't have air temperatures for Ivigtut per se, ice core data for Site J (66o51.9'N/46o15.9'W, 2030m) in South Greenland, inland, was used to reconstruct June temperatures for Jakobshavn (69o13'N,51o6'W,30m) on the west coast. This was -0.11oC in 1633, 0.08 in 1634, 0.30 in 1635, and remained above 0oC for the rest of the 1630s. It was -0.17 in 1854 and was under 0oC from 1857-1875 (Kameda). Ivigtut should have been warmer than Jakobshavn.
In the 1632 universe, we also have a Danish colony on Hudson Bay founded in 1634. We can glean a bit of climate information from the reports of the expeditions that visited Hudson Bay or its southward extension James Bay. Hudson entered Hudson Bay in early July and was frozen into James Bay on November 10, 1610. Luke Fox entered Hudson Bay in late May, 1631. Thomas James entered Hudson Bay on July 16 and James Bay in early September, 1632.
Otherwise, unfortunately, we must decide which part of the post-1700 (Guiot) or 1750 (Catchpole; Ball) data is most analogous to the 1630s and 1640s.
Russia
The wetness/dryness index (norm 10) for Russia (35oE) in the high summer (July-August) is 11.5 for the 1630s and 7.5 for the 1640s. LambCPPF 562). The winter mildness/severity index ( norm zero) was -10 for the 1630s and -36 for the 1640s (the worst value over 1100-1969). (564). Brooks (249) says "In the 1640s . . . severe cold was reported for every month of winter, making this the coldest decade in Russian history since the twelfth century."
North America
For an overview, a good place to start is the North American Drought Atlas, which includes Palmer Drought Severity Index (PDSI) values, for summer 1634-1639, based on tree-ring data (Cook). You can see that the northeast suffered a drought in 1634, which deepened in 1635. California was wet in 1635-6, but it and indeed the entire Pacific Northwest dried up in 1637-39. Mexico was generally quite wet. Unfortunately, rainfall patterns are likely to be perturbed by the RoF. Hence, take Fig. 4, which shows the pattern for 1635-37, with a very large grain of salt.
Figure 4
There is also a tree-ring reconstruction of annual and seasonal temperature and precipitation anomalies for the USA from 1602 on. From this we can see, for example, that in the northeast, both winter and summer 1634 and 1638 were relatively cool, while 1635-37 were characterized by relatively warm summers and relatively cold winters (Fritts).
The 1630s were probably not as bad as 1608; Champlain found bearing ice on the edges of Lake Superior in June 1608 (LambCHMW 230).
Virginia. We don't know for sure why the Lost Colony of Roanoke Island disappeared, but it probably was at least partially attributable to 1587-1589 being the driest three years in eight centuries. The second, successful English attempt to colonize Virginia teetered on the edge of failure. The English colony at Jamestown was founded in 1607, and 1606-1612 was the driest seven years in a 770 year period (reaching a nadir of Palmer Hydrological Drought Index [PHDI] -2.323 in 1610). Despite its coastal position, Tidewater Virginia exhibited droughts in the future, too; there was a short one, for example, in the late 1630s (PHDI of -1.687 in 1637 and -2.67 in 1638) (Stahle).
Spring water temperatures have been reconstructed for Chesapeake Bay based on crustacea mineral ratios. Temperatures were high in the first quarter of the 17th century (higher, in fact, than in the Medieval Warm Period), but then declined, reaching a low in the mid-18th century. In our period, we have 16.6 (1638), 8.97 (1642) and 10.74 oC (1646). The most recent value was 12.5 oC (1995) (Cronin)
New England. In Eastern Massachusetts, the 1630s were cool and wet, while the 1640s were cool and damp. In southern New England, there were two dry growing seasons in each decade, but no years with floods. Note that the same year can be both; the 1780s had six dry years but eight with floods. (Baron).
Eastern Canada. Southwestern Ontario has been identified as having a cool dry climate from 1600 to 1750 (Buhay). At Quebec City, there has been a study of the ice bridge formation (IBF) rate on the Saint Lawrence River. The river at that point is one kilometer wide. Surprisingly, the IBF frequency in 1620-1800 was 16%, less than the 48% seen in 1801-1910. There were seven IBFs in 1620-1660, and none in 1661-1740. Contrast that with the 80% in 1866-1885. Thus, in our period of interest, southern Quebec is relatively warm! (Houle).
Tropical America
In tropical America, the seasons are wet or dry, not hot or cold. The seasonal variation in rainfall is correlated with the seasonal movement of the Inter-Tropical Convergence Zone (ITCZ), northward in the NH summer and southward in the NH winter. In NH summer, it's over northern Venezuela, and that's the rainy season.
In northwest Yucatan and in Venezuela, the general period 1500-1800 was marked by drier conditions, usually attributed to a "southward displacement of the ITCZ and hence reduced trade wind moisture supply to the Caribbean during the summer." However, at Lago Ver
de at the Isthmus of Mexico, a "particularly wet area," this change was apparently not enough to create a moisture deficiency. Indeed, 1600-1650 seems to have been a very good half-century for the tropical forest growth there, suggesting that the dry season was shorter than usual (Lozano-Garcia; Peterson).
On the other hand, in northeast Peru, the LIA was about 10-20% wetter than the 20th century, with the late 16th and the 17th centuries being the wettest period (Reuter).
In the Venezuelan Andes, there were glacial advances in 1180-1350, 1450-1590, 1640-1730, and 1800-1820. (Polissar). However, in the Cariaco Basin (offshore Venezuela), sea surface temperatures were higher in the 17th century than in the 20th (Reuter).
In central Chile, the 1630s were a bit drier than normal (LambCPFF 638), which would have made rivers easier to ford, but also reduced crop yield.
Africa
Information on African conditions is pretty limited. Grove (38) says that southern Africa experienced "sudden warming from 1500 to 1675." In the Makapansgat Valley of South Africa, per speleotherm data, temperatures seem to have been slightly depressed (relative to 1961-1990 base) during the period ~1320-~1750 (Tyson fig. 3). As best I can tell, in the 1630s we were edging up from a low (about 0.5oC below the base) reached around 1600.
In equatorial Africa, the 17th century was marked by droughts in the west and center, and (at least after 1625) wet conditions from Lake Victoria east (Russell; Verschuren). It has been suggested that megadroughts in West Africa were associated with anomalously strong Atlantic trade winds (Overpeck), which would have facilitated colonization of the Americas. The area of Timbuktu was punished by famines due to drought (in 1617-1743) and by great floods (in 1640-1672 and 1703-1738), often in the same year (LambCHMW 226).
Grantville Gazette 37 gg-37 Page 24