November 2010 Archives

Tomorrow I'll lecture on a proxy that Keith mentioned in his presentation - sand dunes. Relic dune fields are one of the main sources of information about the impacts of past droughts on the Great Plains, so I'll review briefly the distribution of dunes in North America, the basics of optical dating and some of recent findings about the behavior of these systems in the 19th century. After that, we'll track back to isotopes, building on our discussion from two weeks ago.

Slides for tomorrow are here.

As promised, I'm only going to assign two papers for discussion next week. They both continue our discussion of stable isotopes from November 17. The first describes spatial patterns of delta 180 in precipitation and shows how latitude, altitude and moisture transport patterns create a 'global fingerprint' in water chemistry. The second paper uses stable isotope records from trees without rings to argue that El Nino was the true culprit responsible for the demise of the Monteverde golden toad.

Spatial distribution of 18O in meteoric precipitation

Tropical cloud forest climate variability and the demise of the Monteverde golden toad

Courtesy of Lane J:

A Relationship Between Cosmic Radiation and Tree Rings
Dengel et al., New Phytologist, 2009

Sigrid Dengel and company performed a tree ring analysis of Sitka Spruce cross sections taken from a single tree plantation in Scotland, UK. The collection's annual rings were measured and correlated with numerous climatic and atmospheric factors to determine relational significance. Of eight factors including temperature, precipitation and cloud base, cosmic ray flux was the factor found to have positive statistical correlation to tree ring growth.

Based on the findings, Dengel et al. suggest that galactic cosmic radiation (GCR) could be used to explain tree-ring growth patterns across the globe. The researchers propose that high galactic cosmic radiation increases aerosol particles in the atmosphere. These aerosols scatter incoming solar radiation creating diffuse solar energy that supposedly stimulates photosynthesis. A more direct approach where GCR directly affects tree growth is also given as a possibility. The class was skeptical.

The article has received negative reviews from portions of the scientific community for a number of reasons including: poor statistical methods, limited information presented on statistical methods, choice of study area, age of samples, and study extent. In contrast, climate change skeptics have used the study to cast doubt on the accuracy of dendroclimatological studies in the prediction of past climate. This is another challenge scientists will have to overcome in getting their climate change message across.

Time Scales of Climate Change
Bartlein et al., Encyclopedia of Quaternary Science, 2006

We took a break from focused research to get an overview by PJ Bartlein on the character of climate variations over time. Scott was kind enough to break down the terminology used within the paper, much of which can be reviewed in the survival guide posted November 9th.

Spurious periodicity (SP) was a hot topic, especially for Keith. Random data without any cycles or patterns can be interpreted to have meaning when certain statistical methods are applied. Any interpretations made from patterns discovered during dataset analysis should be linked with an obvious mechanism to account for SP. It was noted that humans love to assign meaning to random chance.

In general, Bartlein made it clear that climate oscillations are found across a wide range of time scales. Across these time scales, oscillations can be broken up into periodic and quasi-periodic patterns. External climate forces like the Milankovitch cycle or the earth's annual orbit can be predicted with accuracy and are considered cyclical. Quasi-periodic oscillations like ENSO have a distinct pattern but cannot be predicted with certainty. These patterns can be compared on a variance spectrum (see fig. 2) to show differences in frequency and magnitude. Climate variability occurs at all levels of time, but variability increases with scale.

Time series can be expressed by frequency. This was discussed by Bartlein but expanded upon by Scott. With spectral analysis, different colored series represent a specific level of variance and periodicity - red series being slow changes (orbital cycles), blue being fast (ENSO), and the white series as steady variation without a discernable frequency (precipitation).

We have a busy schedule tomorrow: A stable isotope mini-lecture, a presentation by Xiuju and the hazards papers. My slides for the lecture are here.

MAYOR: What do you mean, biblical?
RAY: What he means is Old Testament biblical, Mr. Mayor. Real wrath-of-God-type stuff. Fire and brimstone coming from the sky! Rivers and seas boiling!
EGON: Forty years of darkness! Earthquakes! Volcanoes!
WINSTON: The dead rising from the grave!
PETER: Human sacrifice, dogs and cats living together, mass hysteria!

Next week, I'll spend the first part of class giving a short lecture on the use of isotopes in paleoclimatology. After that, we'll discuss three papers that examine how proxy evidence can be used to understand the dynamics of forest fires, earthquakes and hurricanes.

Tree-ring evidence for an AD 1700 Cascadia earthquake in Washington and northern Oregon

Lake-sediment record of late Holocene hurricane activities from coastal Alabama

Fire-southern oscillation relations in the southwestern United States

Thanks to Ming C. for preparing a very thourough summary of our discussion from last week, including the opening lecture on the past history of Red River floods.

This week we learned about an extreme hydrological event- the Red River floods: the reason why it is so severe, how often it happened in the history and how to prevent it from happening again. One reason that Red River floods are so severe is the geology of the Red River valley. From the Digital Elevation Model of USGS we saw that the Red River Valley is flat. That was caused by the deglaciation of North America. A huge Glacial Lake, Agassiz, formed in the Red River Valley in early Holocene during the deglaciation. With such a flat basin, nothing can hold the water when the flood comes. To study the frequency of the Red River floods in history, we can look at tree rings. However, in 1800s, trees on the river bank were completely cut down for construction. The live trees record can only go back to early 1900s. To track back further, timbers from historical buildings and sub-fossil logs from river alluvium are used. Earlywood vessels, which appear circular in cross-section, are formed during spring growth and used for vertical conduction of water and other nutrients. Prolonged inundation during spring flooding causes shrunken vessels. The flood-ring record goes back 352 years to 1648 and suggests that the Red River flood of 1826 was the largest. To prevent the Red River from severe flooding again, the city Winnipeg started to expand the floodway.

Aztec drought and the "Curse of One Rabbit"
Therrell et al., BAMS, 2004:

This interesting paper was developed by two paleoclimatologists and an archaeologist to evaluate the Aztec belief in cyclical drought-induced famines associated with the calendar icon One Rabbit.

There were 13 drought events recorded in major Aztec historical annals between 1331 and 1543. These recorded Aztec drought years are used to compare with tree ring proxies from Northern Mexico. Results of superposed epoch of analysis shows the mean ring width index for the 13 Aztec drought years is 0.86. However, there were 4 years that the tree ring width is greater than 1. The correlation between the tree-ring values and the Aztec drought years is not as strong as what we'd expected. Significantly above-normal growth occurred 4-year prior to the Aztec drought years. This reminds us of ENSO, which has a strong influence on modern climate over portions of Mexico.

According to the Aztec calendar system, One Rabbit years occur every 52 years. The superposed epoch analysis was taken out again for 13 One Rabbit years between 882 and 1558. The mean ring width index was just above the long-term average during these One Rabbit years. The curse of One Rabbit appears to have been purely coincidental.

Climate and the collapse of the Mayan Civilization
Haug et al., Science, 2003:

The Maya civilization developed in a seasonal desert. Consistent rainfall cycle was essential to support agricultural production. Seasonal variations in the mean position of the Intertropical Convergence Zone (ITCZ) over Mesoamerica and northern South America control the pattern and timing of rainfall in this region. This paper used the bulk titanium content of sediment in the anoxic Cariaco Basin to study the hydrological cycle over northern tropical South America. Further, to infer the relation between the collapse of Maya Civilization and the change of regional climate.

The Cariaco Basin was chosen because: 1) the sediment there is rapidly deposited (30cm per thousand years); 2) The oxygen level is low in the water. The Cariaco Basin is a "dead zone", where no insects or animals act there. This makes the undisturbed sediment in the Cariaco Basion of great value. 3) Paired annual laminations in the sediments reflect large changes in rainfall and wind that occur in this region in response to seasonal shifts in the position of the ITCZ.

The way that this paper uses ITCZ to connect southern Mexico and Cariaco together is still under discussion. The conclusion of this paper is, within the uncertainty of the age model, the proposed end dates (of the Maya civilization) match the three most severe drought events inferred from Cariaco Basin record.

The bigger picture
Caseldine and Turney, JQS, 2010:

This paper discussed three society-specific scientific issues:

• Relevant climate events need to be accurately and precisely dated. It is possible to match things accidentally in paleoclimate studies. Among other proxies, tree ring data and ice cores are the most reliable sources.


• Because there has in the past been a tendency to look for 'events' or catastrophes, little attention has been paid to stable periods or 'normal' conditions. That adds difficulty to estimating severity of short period events.


• How to reconcile differences between proxy reconstructions and calibrate them remains an issue.

Interdisciplinarity will become more and more important in the future. How palaeoscientists respond to this is an open question.

Tomorrow we walk through a 'survival guide' for time series analysis in paleoclimatology. The graphics for our discussion are here.

The set of readings for next week is a little shorter than usual (only 2 papers!) but deals with several issues that are often important in paleoclimate (and climate studies).

The first article, written by the University of Oregon's Patrick Bartlein, is another selection from 2006's Encyclopedia of Quaternary Science. Bartlein explains that the climate system changes at preferred timescales, which are often linked to physical mechanisms. He also shows how certain types of analysis can sometimes produce spurious or false results.
Bartlein Encyclopedia of Quatenary Science 2006.pdf

We'll also read a research paper that argues cosmic rays have a detectable influence on tree growth in Scotland:
Dengel New Phytologist 2009.pdf

This suggestion has been received skeptically by some people in the paleo community. Probably the most detailed critique I've seen appears on a blog written by a pair of anonymous climate scientists. Check it out:
Seeing Red (Noise): Galactic Cosmic Ray Fluxes and Tree Rings

The set of slides for this morning's discussion includes (1) tips on making better professional presentations, (2) the natural history of Red River floods and (3) a few illustrations related to our discussions of the paleoclimate-society papers.

http://www.slideshare.net/scottstgeorge/class-8-past-climates-and-society

In a complete coincidence, my blogroll included an entry by Matt Welsh, a Computer Science prof at Harvard on Conference Talk Pet Peeves. Don't read your slides, don't have ugly graphics, try to be interesting - sounds like solid advice to me.

This week's summary comes to us courtesy of Lu H.

Lewis EOS 2008 paper told us an amazing story about the water levels of Great Lakes. This paper started with a description of modern hydrologic situation of the Laurentian Great Lakes, giving us a vertical scale of the water level change of Great Lakes: in the past century, water levels in all lakes has varies by less than 2.1 meters; compared to more than 20 meters dropped 7~8 thousands years ago causing hydrologically closed lakes. Then the paper shows the evidence for these lowstands, such as submerged tree stumps, and small lakes uplifted and isolated from larger water bodies, low-level erosion , microfossil evidence of brackish conditions, a pause in outlet basin sedimentation, and submerged beaches. Later, the paper explained the lowstands in the Great Lake Basin which is a dry climate substantially drier than the present climate. The authors quoted studies from hydrologic modeling of present lakes showing that in order to achieve lake closure, mean annual precipitation would have to decrease by about 25% - 42%, and mean temperature would have to increase 5 degree C.

Laird PNAS 2003 paper is a story about large-scale moisture shifts across the northern prairies of North America during the past 2000 years. Authors using 6 lake sediments across the North American prairies show that two different types of major shifts in moisture regimes (from wet to dry or from dry to wet) existed for at least the last two millennia. The evidence they used are diatom analyses from lake sediment cores: diatom-inferred salinity estimates (indicator for effective moisture), total phosphorus inferred from diatoms (indicator for the degree of water-column mixing and stratification as related to general climatic conditions). Such large-scale moisture shifts were thought to be caused by the expansion of the polar vortex, and may be a common phenomenon which may reoccur over longer time frames. The differences in timing of moisture shifts were explained as the fluctuations in the position and shape of the jet stream.

Shanahan Science 2009 paper is one of the most interesting papers we have discussed (at least I think) although we didn't have too much time to talk about it. This paper combined geomorphic, isotopic, and geochemical evidence from sediments of Lake Bosumtwi, Ghana to reconstruct natural variability in the African monsoon over the past 3000 years, and to link the severe droughts to Atlantic temperature variations. The big motivation of such research is that current severe drought happened in west Africa in recent decades , and study past climate may have some indications for future climate change (of course this is also the motivation of almost of all paleoclimate research--- study past to shape future). The main evidences in this paper are 1. Oxygen isotopes of authigenic lake carbonate (indicator for precipitation) 2. Elemental concentrations of sediment (indicator for size of the erodible catchment area --- high or low lake levels). By using spectral analysis, the authors were able to identify the several variabilities of the Atlantic multidecadal oscillation (AMO). By examining the linkage between the AMO and west African monsoon (WAM), the paper concluded that the WAM has also varied on multidecadal time scales. This conclusion suggested that more severe than current droughts will eventually come back.