Historic variations in temperature number Four-The Hockey stick

Guest essay by Tony Brown

Section 1 Summary of a previous article;

A short while ago I published an article on ‘Noticeable climate change’ during the past 500 years, based on historical observation and instrumental records. To understand the context of this current article –the purpose of which is to extend and amplify this earlier work- it is linked to below and a brief recap of its findings has been made in the following paragraphs.

http://judithcurry.com/2013/06/26/noticeable-climate-change/

The referenced study noted that our climate changes frequently when calculated on an annual and decadal basis, in fact virtually no decade is like its predecessor or successor. Sometimes the change is fairly small but is often so ‘noticeable’ that humanity and nature will be affected. Figure 1 below illustrates this effect, where the brown verticals represent annual temperatures, the green is decadal and the red line represents fifty year steps.

Figure 1; Annual, Decadal and 50 year periods from CET

The analysis is based on Central England Temperature (CET) to 1659 which is the world’s longest instrumental record, and my own reconstruction from that date to 1538. CET is said by a number of scientists to be a reasonable (but by no means perfect) proxy for global and to a greater extent Northern Hemispheric temperatures. In considering the historic instrumental temperature record and the paleo proxy reconstructions (shown in figure 2) that cover much longer periods, we should heed Hubert Lamb’s maxim that ‘We can understand the (temperature) tendency but not the precision’ so whilst accuracy to tenths of a degree is impossible, determining the general shape and direction of temperature travel is reasonable. Certainty in determining great historical temperature precision lies only in models.

In then providing figure 2 below we can observe a number of results that can most usefully be summed up by repeating the conclusions from the original article;

Figure 2-paleo reconstructions plus CET in annual, decadal and 50 year steps

These conclusions were;

*By any criteria, climate change is frequent, indeed the ‘norm’. Variability of our climate on a decadal basis is considerable and is even greater on an annual basis.

*These decadal episodes of variability appear greater in the past than in modern times as can be seen in the sharp drops, then recovery, during the LIA episodes.

*However, there are tight 50 year ‘paleo’ boundaries, with the frequent oscillations currently on a slightly rising trend from the start of the instrumental temperature record in 1659, albeit with a sharp reversal in the last decade.

*There are some hints of a similar rise to broadly equivalent modern levels around 1540 in the reconstructed CET.

*Rising temperatures reflect the relative dearth of ‘old fashioned’ winters in the second half of the record, as well as changes in other seasons. Summers have the tightest boundaries, appreciably lower than the other seasons.

*Humans need to make substantial accommodation to cope with even relatively short periods of ‘different to normal’ weather. Consequently it might be useful for the authorities to consider ‘noticeable’ climate change as covering decadal rather than 30 year periods

*Paleoclimate reconstructions (figure 2) capture long term climate variability of 50 years and more quite well, but fail to pick up the much more variable fine grain of annual and decadal variability. Consequently a false picture of apparent climate stability in the past is portrayed.

* As paleoclimate reconstructions are usually only measured against global instrumental temperature records commencing 1880 they do not find any of the ‘hockey stick’ effects that can be seen in the older instrumental temperature records.

*When instrumental records are not available, the historic observed record of the past millennium is likely to be a valuable aid to the development of paleoclimate reconstructions using proxies.

Section 2 An examination of the ‘hockey stick.’

The so called ‘spaghetti’ graphs used in figure 2 above are interesting, but their range of variability (excepting Moberg et al) remain almost as limited as their iconic predecessor the ‘Hockey stick’ produced by Dr Michael Mann et al and from which the IPCC third assessment report graphic from 2001 was derived.

http://en.wikipedia.org/wiki/Hockey_stick_graph

Consequently in view of its continued importance and to demonstrate the strengths and shortcomings of his model, and many of those that followed, it is reproduced below (red line) together with Hubert Lamb’s early pre computer graphic (blue line) that was the inspiration for the IPCC’s global temperature chart used in their first 1990 assessment. This is overlaid on to the CET data already referenced: light blue line 50 year segments and brown line for decadal information. It is all rebased to zero anomaly. The yellow shading illustrates the significant differences between the Mann and Lamb reconstructions and graphically illustrates the 1000 year long period of climate stability depicted by Dr Mann as compared to the much more variable climate researched by Lamb.

Figure 3

In order to be able to make an easy visual comparison, in the next graph we have removed the 1990 IPCC graph and rebased CET, so the anomalies match with Dr Mann’s reconstruction which –until the modern instrumental record is inserted at the end- runs between minus 0.2 and minus 0.5C anomaly-somewhat cooler than CET.

Figure 4 CET rebased with Mann et al 1998

The difference between a warm period such as the 1730 decade at plus 0.4C and a cold decade such as 1690 at some minus 1.4C anomaly measured via instruments can be seen to be far greater than the variations that any of the paleo reconstructions in figure 2 pick up . The Mann et al Hockey Stick (Figure 4) also illustrates very limited variability throughout its 1000 year long proxy reconstruction. The relatively short and intense perturbations noted above since 1538 are however again well captured in CET instrumental records and observations.

Section 3 Pros and cons of Paleo temperature proxy reconstructions.

Such proxy material as tree rings cannot be as accurate as instrumental records or detailed reconstructions using a variety of observational material-but there are nevertheless a number of obvious consequences that those who debate climate as either ‘realists’ or ‘sceptics’ need to face when considering this data;

*The first is that Dr Mann’s graphic (as do many of the other paleos) make a pretty god job of picking up the relatively limited temperature variability we can observe over a 40/50 year or longer period . This is confirmed when comparing the data with the CET 50 year instrumental ‘paleo’ (the horizontal blue line.) This is with the notable exception of the coldest period of the Little Ice age around 1690 and the subsequent recovery in the following decades, providing the most notable hockey stick in the record.

*The CET comparison to the global instrumental record (shown from 1900 in figure 4) is pretty good as is its comparison to the 50 year paleo records. Britain as a temperate country will have different climate characteristics/variability than the tropics or countries at other latitudes but it can be seen that CET provides a useful long term record validation, although Lamb’s maxim should be borne firmly in mind and precise accuracy and correlation at all times is impossible.

*The variability shown in the uptick from 1900 looks unusual only because an instrumental temperature record-which captures variability-is now used, whereas the long term paleo reconstruction proxies previously used, do not have this ability to capture short term variability and thereby present an impression of a ‘stable’ climate. The uptick is therefore purely an artefact of changes in methodology as a ‘paleo’ apple is swapped for an ‘instrumental’ orange.

*The 40/50 year paleo reconstructions (figures 2, 3 and 4 ) fail to capture the decadal variability (orange lines) let alone the annual range (shown in figure 2 as a brown vertical line). They consequently fail to ‘see’ such notable events as the great warming centred on 1730, the recovery around 1830 from the coldest decade (1810) since the depths of the LIA in 1690, and the final bursts of the LIA in 1840 and 1890.  Looking further back, the paleo reconstructions also do not replicate the considerable drop to the depths of the LIA around 1690, the (reconstructed) warmth around 1630, the period of well documented cold at the beginning of the 17^th century and the sharp (reconstructed) rise around 1540 to something apparently approaching the temperatures at the end of the 20^th century. In particular the paleo proxy reconstructions represent the severe perturbations of the various periods of the Little Ice age as merely shallow downwards blips, whilst the astonishing recovery around 1690 featuring the largest hockey stick in the record is a corresponding shallow upwards blip

*if the paleo proxy reconstruction can miss these considerable perturbation downwards, some doubt is thereby introduced as to whether they would catch other similar perturbations in the opposite direction most notably during the so called medieval warm period.

*it confirms that the instrumental temperature record shows an upward trend (with various reverses and advances) from the start of the CET instrumental record in 1659 making the 1880 start point for the instrumental global record used by GISS appear to be merely a staging post in the upwards trend, rather than the starting post.

*The sharp downwards trend in Britain from 2000 currently causing such concern to the Met Office can be seen in historical context as merely another episode of ‘noticeable climate change’ readily captured in the instrumental record, but this time that of cooling rather than warming.

*At around the year 2000 the real world CET data diverges from the composite global temperature comprising of tens of thousands of averaged and smoothed records. Amongst these are those showing cooling, warming and stasis which seem to be roughly cancelling each other out to create a ‘pause‘ in warming that is currently the subject of much debate.

The very limited deviation from the considerable climate stability illustrated throughout the paleo reconstructions -including the Mann et al 1998 ‘hockey stick’- is difficult to corroborate with actual instrumental or historic evidence. Similarly the continual downwards trend in temperature from the start of the Mann et al records from 1000AD to 1900 does not seem to be validated by other data.

Section 4 Comparison of paleo to Glacier movements

In this final graph we have calculated historic glacier movements during the last 3000 years. (See note 1) Over it we have inserted the Mann et al 1998 data covering the past 1000 years together with the decadal record from CET back to 1538.

Fig 5-3000 year Glacier movements with CET decadal/50 year steps and Mann et al 1998

A closed blue horizontal line at the top of the graph equates to a period of glacial retreat (warmth) and a closed blue line at foot of graph demonstrates glacier advance (cold)

That glacial movements can be surprisingly short lived can be seen in the century long glacier advance around 1200 to 1300 AD, and to a lesser extent the 30 year retreat around 1730. Such short changes as noted in this latter period may be relatively common, but the records are unlikely to exist to be able to trace them in earlier times.

The small temperature deviations from the ‘norm’ shown in paleo proxy reconstructions- including that of Mann et al 1998-seem most unlikely to be of a scale that can precipitate glacier movements of any consequence. Several consecutive warm cold decades that can be noted in the instrumental records will however likely start such movements which will be accentuated if the prevailing characteristic of warmth or cold lasts for some time. In the case of the MWP this period of warmth lasted around 450 years . (Clearly however brief Warm periods can occur during a general glacial retreat and brief cold periods during glacial advance.)

That the paleo reconstructions somewhat accurately capture long term variability makes this feature useful. However, they appear to comprise of a very coarse sieve that allows the real world of constant noticeable climate change with considerable temperature swings that affects us all to slip through unnoticed. This makes the use of paleo reconstructions as the basis for far reaching policy changes somewhat problematic and counter intuitive as it is based on a belief that the past comprised of a relatively unchanging climate. A belief that is contradicted when the real world annual and decadal record is closely examined..

That CET appears to be a good –but not perfect- proxy of global temperatures can be seen in the paleo and instrumental record. There is a considerable body of literature from a number of leading climate scientists and related organisations that suggest that CET appears to be a reasonable but by no means perfect proxy for temperatures far beyond the shores of England. The author is preparing a piece for peer review entitled; ‘ Is CET indicative of Global or Northern Hemispheric temperatures?’

A future article will concentrate on the far greater extremes that can be noted in our historic weather events than in the modern record, perhaps not surprising in view of its observed greater variability and considerable historic perturbations.

A future article will also delve further back from the CET reconstruction to 1538 detailed in Part 1 of ‘The long slow thaw?’

http://judithcurry.com/2011/12/01/the-long-slow-thaw/

Early analysis suggests some evidence of a period around 1500 around as warm as 2000 and a sharp, but as yet not fully researched deterioration in the few decades prior to that which appears to have some corroboration in the glacier data which was researched separately. Beyond that period work continues in assembling the necessary historic and scientific material to enable the continuation of the reconstruction with some worthwhile degree of accuracy .

In looking at the disparity between paleo records and instrumental observed records It would be useful to see historical climatologists and modellers work more closely with each other in order that the past climate states and their variability can be more accurately depicted.

References and Notes.

Note 1 Glacier records have been painstakingly researched by numerous glaciologists and historians over many years. Their sources include church records, commissions of inquiries into glacier disasters, taxes on farms affected by glaciers, town records, population records, illustrations and lithographs, observations by travellers and scientists, scientific papers, historic articles on glaciers from contemporary sources in English, French, German and Italian, correlation with wine and grain harvest dates, alpine clubs, mountaineers and tree line/plant growth records amongst other sources. See references under.

References

‘The Little Ice Age’ by Professor Brian Fagan

‘History and climate’ edited by P D jones

‘Little ice ages ancient and modern Volume 1 and Volume 2’ Jean Grove

‘Climate history past and present’ Hubert Lamb

‘Times of feast, times of famine –a history of climate since the year 1000’ E Le Roy Ladurie

‘Paleoclimatology, reconstructing climates of the quaternary’ Raymond Bradley

‘Little Ice Age’ Michael E Mann Volume 1, The Earth system: physical and chemical dimensions of global environmental change,

http://www.meteo.psu.edu/holocene/public_html/shared/articles/littleiceage.pdf

P154 on Groves, Jones, Matthews research.

http://books.google.co.uk/books?id=-cgnFsLkIAYC&pg=PA157&lpg=PA157&dq=glacial+advance+1220&source=bl&ots=y1cKpnhMdB&sig=hhX8s1K9KtGjIukXIepY2HSgC3M&hl=en&sa=X&ei=I7LmUZnGCKmh0QWnxIGQBA&ved=0CGUQ6AEwCA#v=onepage&q=glacial%20advance%201220&f=false

Barclay et al

http://web.cortland.edu/barclayd/publications/2009_QuatRes.pdf

Glaciers around 1640; fragments of reports are available for the period, such as the 1955 study by Guichonnet which suggested three Chamonix glaciers had reached a maximum by 1640 then retreated by small amounts. No doubt other detailed information will become available as more archival material is discovered.

The World glacier monitoring service records the more recent history of glaciation mostly over the last 150 years

http://www.grid.unep.ch/glaciers/pdfs/glaciers.pdf