Putting aside the question of whether the Medieval Warm Period was a global or regional event, I wanted to touch on the Norse colonies in Greenland. Greenland was explored and named by Erik the Red between 982 and 985. Most accounts indicate he used the name Greenland to lure settlers from Iceland to join him at this far-flung outpost. Erik and found two areas in the southwest suitable for settlement, with grasslands and small stands of alder and birch. Most of the rest of the island was covered with arctic tundra, snow and ice, much as it is now. Some of Greenland's ice is over three kilometers deep. The first group of settlers - 24 boatloads - set out from Iceland in the summer of 986. Only 14 ships finished the journey, with the other 10 forced to turn back or lost at sea.

The Norse colonies grew and prospered for at least the first 200 years in Greenland, with the population peaking at about 5000. Ships from Norway and Iceland visited once or twice a year, and the colonists traded live falcons, polar bear skins, narwahl tusk, walrus ivory and hides for timber, iron, tools and even some luxuries. The Greenlanders had cows, sheep and goats which grazed when the weather permitted but had to be housed in barns during the harsh winter months. As the Medieval Warm Period gave way to the Little Ice Age, Greenlanders kept their livestock in their homes for the added warmth. This was happening as early as the 12th century.

As the years passed, the farmland lost its fertility and the cutting of trees for fuel and for the production of charcoal lead to erosion. The Norsemen were stubborn, proud people. They saw themselves as European and continued to pursue a European lifestyle. The Inuit, who also were living in Greenland in the Middle Ages, adapted well to the changing climate. The Norsemen refused to adopt the clothing and hunting gear which allowed the Inuit to thrive, believing the European way to be superior.

The last written record from the Greenland Norse colonies dates from 1408. A few Norsemen remained in Greenland into the 16th century.

It's fairly clear the Greenlanders lived a hard scrabble life. They were accustomed to hardship. A program on the Little Ice Age from the History Channel mentioned that even during the period of climate optimum, most Europeans had few resources. A failed crop lead to famine. The same must have been true in Greenland. Climate change is not believed to be the sole cause of their demise, but was probably the primary stressor. They had little economic "cushion" to fall back on.

I doubt that paleoclimatology is ever taught at the 100-level in college; however, if it were, then Dr. Richard Alley, Evan Pugh Professor of Geosciences at Penn State, would be one of the best instructors to have. Dr. Alley, who studies paleoclimates from ice cores and who is one of the authors of the IPCC's 4th Assessment, has kindly taken time to answer some questions for the blog. Because of length, I will have additional answers in future posts.

Question: Dr. Alley, you study ice cores. Other paleoclimatologists study tree rings, coral, fossil records and sediment cores. Are all those sources taken into account when creating a climate record, or would a climate reconstruction be produced from each source?

Answer: Yes, and yes.

Typically, a researcher or group of researchers will develop a climate record based on one or a few cores. This most typical is a local climate record, giving for example the temperature at the site where the core was collected. (As you might imagine, there are careful analyses that go into such a statement--if you are relating pollen to temperature, then the temperature you infer from the pollen present represents conditions over the region from which the pollen blew to your core; if you are studying indications of temperature in an ice core, then the motion of ice--glaciers flow--may have brought samples from somewhat higher elevation. These sorts of complications are generally handled in the technical literature, scientists love to argue about them, and so we probably do a pretty good job with them.)

Once many such records are developed, they are pieced together to reconstruct regional or global temperature. Some selectivity is often used in this. For example, in the studies of the most recent millennium, some workers have restricted themselves to annually resolved records (those in which the age is really known to the year). This is valuable, in that it avoids any problems with estimation of ages (is this wiggle in my core really the same climate event as that wiggle in your core?), but it loses information from cores that, say, have 10-year dating accuracy. So other groups have assembled lower-time-resolution histories from lower-time-resolution archives.

This is a little more technical, but.... As you go older, annually resolved records become rarer and rarer, so we use other approaches. For example, in an ice core from Greenland, we obtain indications of the temperature in Greenland, but we also see changes in the dust of the atmosphere. People have carefully looked at the isotopic, chemical, etc. composition of the dust, and it came from Asia (from the high plateaus of central Asia, blowing to the high ice sheet of Greenland). There are HUGE changes in the dust (10-100 fold), and explanations of the dust changes all invoke changes in the source in Asia (you could change how much falls out on the way, for example, but it is hard to find a reasonable model that "works" to give the very large Greenland changes). So, we infer that there were changes in the Asian dust source, probably linked to wet/dry (with more dust when the rains stop). So, people go to Asia, and they develop records of wet and dry, and find that the record of wet/dry in Asia looks like the record of dusty/not in Greenland ice, agreeing in size of changes, duration of changes, and in age of changes, within the uncertainties in knowing exact ages. So we then believe we can "tweak" the age estimates, within the known uncertainties, to improve the correlation.

Here is additional information from Dr. Richard Alley, Evan Pugh Professor of Geosciences at Penn State, talking about climate proxy.

Question: Dr. Alley, is any one climate proxy more accurate than the others? Does any provide more information than the others?

Dr. Alley's Answer: Consider the width of a tree ring. It records how "happy" a tree is by how much the tree grew. Go to a very dry place, and a "happy" tree is one getting rain, so the tree rings are rain gauges. Go to a very cold place, and a happy tree is a warm one, so the tree rings are thermometers. But, the tree will also notice if it is being eaten by beetles, or shaded by neighbors, or other things. So the history of thickness of tree rings from one tree is unlikely to be a very good climate record. But, use a lot of trees, and you start to average over the changes in beetles and neighboring trees, and to see the climate signal. With enough trees, and enough cleverness, you start to learn about temperature and rainfall, over large areas.

Consider the isotopic composition of nitrogen and argon in ice-core bubbles. When snow falls on an ice sheet, the weight of the new snow squeezes the old snow until bubbles are pinched off, trapping air samples. Usually, the snow gets a couple of hundred feet deep, so the bubbles are being trapped way down there. The spaces in the snow are interconnected from the surface down to a couple of hundred feet, but the spaces are small enough that the wind doesn't mix the air, which instead mixes by diffusive processes.

If you warm or cool the surface, it takes a century or so for the temperature to change down where the bubbles are trapped (it takes a short while to burn the thin skin on your finger if you touch a hot burner, much longer to cook a hamburger, and much much much longer to cook a turkey; roughly, make something twice as big and heating it takes four times as long, so the couple of hundred feet to the bubble-trapping depth does take a century or so).

During that century, the temperature at the top and bottom of the snow will be different. When communicating gases not mixed by the wind have a temperature difference imposed, the gases separate (by a TINY amount) with the heavier ones on the cold end, in a process known as thermal diffusion, which is very well understood by physicists. This makes the trapped gases slightly different from the free atmosphere. If one measures characteristics (such as the isotopic composition of nitrogen and argon) that cannot change rapidly in the atmosphere (because there is so much nitrogen and argon up there, and such small gains and losses), then you can learn the deviations, and thus the temperature differences across the snow, and thus the history of temperature change at the surface. This is almost entirely a history of temperature change in the near-surface (there is a "tweak" because the gases also separate a tiny bit under gravity, and the thickness of the snow hence the gravitational effect may change a bit, but by measuring argon and nitrogen isotopes, the temperature and gravity effects can be separated because they affect nitrogen and argon differently. Then, because the gravity depends on the snow thickness which depends on temperature and snowfall rate, you can learn something about snowfall, too.). This is a much easier thermometer than a tree ring, because the gases don't worry about neighboring trees or beetles (which are notably absent on the ice sheets). But, we don't happen to have an ice sheet in Kansas, so we have to use tree rings or something else there rather than ice-core gases.

In short, nature gives us lots of indicators, the indicators are not as easy as going out and reading a thermometer, but we can read the indicators with a bit of effort and care.

As I mentioned last week, Dr. Richard Alley, Evan Pugh Professor of Geosciences at Penn State, has taken the time to answer some of my questions about paleoclimatology. In today's post, he talks more specifically about the information that can be gleaned from an ice core.

Question: Specifically related to ice cores, can you explain what information is captured in glacial ice and how you "read" it? How can you tell the difference between a cold year and a year with higher than average precipitation, for instance?

Dr. Alley's Answer: Temperature, I explained earlier (Trees and Ice are Links to Past Climate).

Another we use (a "fuzzy" one--doesn't reveal short-lived temperature changes a long time ago, but is reliable) is the physical temperature of the ice. If you throw a frozen turkey in the oven, turn on the heat and then leave, and your significant other comes home and wants to know how long the turkey has been in but doesn't have your cell phone number, why, drill a hole in the turkey, measure the temperature, and the colder the inside, the shorter time the turkey has been cooking. The Greenland ice sheet a mile down is colder than the surface is, and is colder than the bedrock two miles down, because the ice is still warming from the cold of the ice age, and we can estimate how cold the ice age was in central Greenland from this.

Another is related to the isotopic composition of the water in the ice. If you have a bunch of water molecules, a few of them will have an extra neutron or two ("heavy water"). The heavier water is still water, but just a tad heavy. The heavier stuff rains or snows out of a cloud first. As an air mass moves over an ice sheet and cools, the heavy falls out, so the remaining vapor comes closer and closer to being all "light", so the next rain or snow becomes more and more "all light". The colder the air mass, the lighter the isotopic makeup of the rain or snow. So, the isotopic history in an ice core is a temperature history. If you look at isotopes, and isotopes of gases, and borehole temperatures, you can start to get a reliable history of temperature at the site.

In favorable ice cores, we date by counting annual layers (summer snow and winter snow look different, are isotopically different, chemically different, electrically different, etc.). We check as far back as we can using known time markers (find the fallout from historically dated volcanic eruptions, or from atomic-bomb explosions), and it works really well. We have several people count several things several times, try really hard not to "cheat" by checking on the answers of others, and then compare our ages to those for correlative climate events in other records (tree rings, etc. from elsewhere). The thickness of an annual layer (after a correction for thinning from compaction under weight of snow above and from ice flow) gives the snow accumulation rate. Changes in "dirtiness" of ice (how much dust) either represent changes in the dust delivery, or in the delivery of water to dilute the dust; know the accumulation rate, and you have the dilution.

A refrozen-meltwater layer shows it was warm. Fallout of odd nuclides made by cosmic rays in the atmosphere tell how many cosmic rays were getting in (and because the magnetic field and the sun's activity help protect us from cosmic rays, tell about the magnetic field and the sun's activity). One can find micrometeorites, pollen, sea salt, etc. and learn about those. And, the trapped bubbles are our only reliable samples of old air and the greenhouse gases in that old air. So, there is plenty to learn.

This week's Headline: Earth video is part 2 of Katie Fehlinger's interview with AccuWeather.com's Ken Reeves, discussing climate ideas from 30 years ago.

Here is our final question-and-answer session with Dr. Richard Alley, Evan Pugh Professor of Geosciences at Penn State. To review, Dr. Alley studies paleoclimates from ice cores and is one of the authors of the IPCC's 4th Assessment.

I'd like to say thank to you to Dr. Alley for sharing so much information with the blog over the last couple of weeks.

Question: Dr. Alley, how can you tell what forcings were involved with past climate changes?

Dr. Alley's Answer: One can use strength of magnetization of lava flows and sediments to learn about the strength of the magnetic field, and then the changes in cosmic-ray-produced things not explained by the magnetic field are probably changes in the sun (we don't think the total cosmic ray flux changes much). There are small changes in the sun recorded, and small changes in climate going along, with a behavior that matches what we expect. The large changes in magnetic field that have happened don't seem to have had an effect on climate, so that doesn't seem to be an important forcing. There is little evidence of changes in micrometeorites/space dust (except for the occasional giant meteorite such as the one way back that killed the dinosaurs), so space dust doesn't seem important. Volcanoes block the sun and bring cold--a degree or two for a year or two--this is well-recorded, but isn't "organized"--one eruption doesn't trigger another, so the volcanoes mostly make "noise". The very skinny version is that climate has been mostly controlled by sun (with small changes in total brightness giving small climate changes), and by greenhouse gases (mostly CO2). In addition, features of Earth's orbit (which move the sunshine around over the planet over tens of thousands of years--more in the north or the south, or more at the equator or the poles) have paced ice ages, in part by affecting CO2. Over long times, drifting continents affect where ocean currents go and other such things, which matter to climate.

Question: What degree of confidence do paleoclimatologists have in the climate reconstructions? Does it diminish farther back in time?

Dr. Alley's Answer: Given the huge range of indicators, the degree of confidence goes from pound-on-the-table/this-is-almost-surely-right to this-is-more-likely-than-not-to-be-right/can't-say-much-more. In general, confidence is better more recently.

Question: Can you compare the current warming in the Arctic with the warming of the 1930s and 1940s? What were the climate forcings that brought on that warming?

Dr. Alley's answer: This is still a topic of research, but the recent paper by Johannessen et al., and other work, give good pointers. The most obvious difference between the earlier warming in the Arctic, and the more recent one, is that the earlier one was mostly restricted to the Arctic, and the more-recent one is almost everywhere. The geographic pattern of the earlier one did not look like that expected for greenhouse gases (which had not risen much yet), and the more recent one does look like the greenhouse-gas pattern. Some worker points to a role for sun and volcanoes in the earlier one (occasionally, by accident, brighter sun and fewer volcanoes happen at the same time), together with a "dynamic" component--perhaps a change in the ocean overturning linked to the Atlantic Meridional Oscillation. The climate is a complex-enough beast that glib answers should be viewed with caution--the IPCC or National Academy statements on global warming are NOT glib, but are cautious and carefully reasoned, and must be.

Last month we learned a lot about paleoclimatology, especially about ice core research. A new study out of NASA's Jet Propulsion Laboratory examines a different kind of paleoclimate record.

Researchers studied Egyptian records of annual Nile water levels between 622 and 1470 A.D. at Rawdah Island in Cairo. Water levels were critically important for agriculture and the records are considered to be "highly accurate." These water records were cross referenced with records of auroral activity kept in northern Europe and the Far East. These records are also considered accurate, as people believed auroras were portents of disaster.

Some clear links were found between solar activity and climate variations, in particular an 88- and a 200-year cycle. These links appear to be connected to the North Atlantic Oscillation, a large scale seesaw in atmospheric mass between the subtropical high and the polar low.

Nigel Weiss, Emeritus Professor in Mathematical Astrophysics at the University of Cambridge, spoke last Wednesday at the Royal Astronomical Society's National Astronomy Meeting. He described how solar activity was important in past climate change, but current global warming is driven by human activity.

The variations in global temperature driven by solar variation tend to be on the order of 0.1-0.3 degrees Celsius, while in the past 100 years, global temperatures have increased by almost 1 degree Celsius. We are currently in a period of high solar activity, which would correspond to some warming. At some point that activity will diminish, though scientists have no way to predict exactly when that will happen or how intense the upcoming minima will be.

Will the upcoming minima counterbalance global warming? Don't count on it, says Professor Weiss:

"Although solar activity has an effect on the climate, these changes are small compared to those associated with global warming," he said. "Any global cooling associated with a fall in solar activity would not significantly affect the global warming caused by greenhouse gases."