When you’re trading in puts and calls (or derivatives), it’s important to know the sign of the relationship between the value of the derivative and the market. Short positions will go up in value as the market goes down. And, unfortunately, you don’t get to decide afterwards whether you wanted to be short or long. Proxies in climate can, in a sense, either be “short” or “long”, in the sense that the values of some proxies (e.g. coral dO18) are said to go down with higher temperatures, while the values of other proxies (e.g. ice core dO18) are said to go up with higher temperatures.
One feels that it is not asking too much of paleoclimatologists to know the expected sign of a proxy derivative. Traders would like to decide on the sign of a proxy derivative after the fact, by taking a correlation to market performance, but this luxury is denied to them, as it should be denied to climate scientists.
In Mann et al 2008, there is a truly remarkable example of opportunistic after-the-fact sign selection, which, in addition, beautifully illustrates the concept of spurious regression, a concept that seems to baffle signal mining paleoclimatologists. For this example, we turn to the highly HS-shaped Finnish sediment series of Tiljander et al 2003. (Update: We reported this in a PNAS Comment, to which Mann responded. See continued discussion at Upside Down Mann and the Peer Reviewed Literature.”)
Tiljander et al cored varved sediments from Lake Korttajarvi, Finland, going back through most of the Holocene. In Tiljander et al 2003, they distinguished the amount of mineral and organic matter in each varve. The basis for using mineral and organic matter as climate proxies is set out as follows:
The amounts of inorganic and organic matter, form the basis of the climate interpretations. Periods rich in organic matter indicate favourable climate conditions, when less snow accumulates in winter by diminished precipitation and/or increased thawing, causing weaker spring flow and formation of a thin mineral layer. In addition, a long growing season thickens the organic matter. More severe climate conditions occur with higher winter precipitation, a longer cold period and rapid melting at spring, shown as thicker mineral matter within a varve.
The caption to their Figure 5 reports the following link between X-ray density and their climate mechanism:
High X-ray density corresponds to high amount of mineral matter (light grey value tints in X-ray film) and low X-ray density corresponds to dark grey values caused by a higher proportion of organic matter.
Putting the two paragraphs together: warmer climate favors more organic material and thus a low X-ray density. In order to show warm values at the top of a graph, you need to invert the plot (i.e. you have to pay attention to the sign of your climate derivative.)
In the figure below, on the left, I show an excerpt from their Figure 5 which they show vertically (only the X-ray density is shown here – consult the original paper for the other plots.) The left portion of their Figure 5 shows an organic-rich period in the MWP, about which they say:
An organic rich period from AD 980 to 1250 in the Lake Korttajarvi record is chronologically comparable with the well-known ‘Medieval Warm Period’ (e.g. Lamb 1965; Grove & Switsur 1994; Broecker 2001). The sediment structure changes, less mineral material accumulates on the lake bottom than at any other time in the 3000 years sequence analysed and the sediment is quite organic rich (LOI ~20%). Thus, the winter snow cover must have been negligible, if it existed at all, and spring floods must have been of considerably lower magnitude than during the instrumental period (since AD 1881). According to the scenarios presented by Solantie & Drebs (2001), a 2°C increase in winter temperature would decrease the amount of snow in southern Finland significantly. Under such conditions, winter snow accumulation and intense spring floods would be rare events….
The Lake Korttajarvi record also indicates a climatically more severe period in the 17th century. Two periods, AD 1580–1630 and AD 1650–1710, are marked by an increase in both sedimentation (varve thickness) and mineral matter accumulation (relative Xray density). Also, magnetic susceptibility values are high between AD 1650 and 1710, indicating increasing mineral matter input into the lake.
They cite literature, including Hulden 2001, showing mild conditions in Finland in the MWP.
On the right, I’ve plotted the corresponding data so that “warm” grey values are on the top. I’ve also highlighted the (MWP) period identified as having elevated values of organic matter. If you squint, you can satisfy yourself that the left-hand and right-hand panels are showing the same data.
Fig. 1. Left from Tiljander et al 2003 Figure 5; right – plot of X-ray density (inverted).
Plotted according to the climatic interpretation offered by Tjilander et al, the modern warm period shows as colder than the Little Ice Age, something which makes no sense if this data is to be used as a climate proxy. Tiljander et al provide a plausible interpretation of the “divergence” of the proxy from its climatic interpretation as a result of agricultural and construction disturbance to sediment patterns, actually tying several especially thick varves to ditch and bridge construction:
This recent increase in thickness is due to the clay-rich varves caused by intensive cultivation in the late 20th century. …
There are two exceptionally thick clay-silt layers caused by man. The thick layer of AD 1930 resulted from peat ditching and forest clearance (information from a local farmer in 1999) and the thick layer of AD 1967 originated due to the rebuilding of the bridge in the vicinity of the lake’s southern corner (information from the Finnish Road Administration).
Now let’s see what Mann et al did with this data. All of the 20th century values of varve thickness, X-ray density etc go up like crazy with the agricultural and construction activities as shown below for 2 of the 4 series (the other two are similar). Instead of using the climatic interpretation of the data described by Tiljander et al, Mann correlates the increases in varve thickness and changes in density and color, originating from local construction and farming, to world climate.
Figure 2. Two of 4 versions used in Mann et al 2008
By flipping the data opposite to the interpretation of Tiljander et al, Mann shows the Little Ice Age in Finland as being warmer than the MWP, 100% opposite to the interpretation of the authors and the paleoclimate evidence. The flipping is done because the increase in varve thickness due to construction and agricultural activities is interpreted by Mann et al as a “nonlocal statistical relationship” or “teleconnection” to world climate. Mann:
the EIV approach, which makes use of nonlocal statistical relationships, allowing temperature changes over distant regions to be effectively represented through their covariance with climatic changes recorded by the network
A more convincing example of spurious regression in “peer reviewed” literature will be hard to find. After reading through this, I keep expecting someone to say:
Live from New York, it’s Saturday Night.
H/t to Howard Wiseman: http://www.youtube.com/watch?v=M1owcncKCHg
TILJANDER, MIA, M. SAARNISTO, AEK OJALA, and T. SAARINEN. 2003. A 3000-year palaeoenvironmental record from annually laminated sediment of Lake Korttajarvi, central Finland. Boreas 32, no. 4: 566-577.