By Andy May
In the last post, I discussed the problems comparing modern instrumental global or hemispheric average temperatures to the past. Ocean temperature coverage was sparse and of poor quality prior to 2005. Prior to 1950, land (29% of the surface) measurements were also sparse and of poor quality. Only proxy temperatures are available before thermometers were invented, but, again, these are sparse and poorly calibrated prior to 1600. So how can we compare modern temperatures to the distant past? We can’t do it globally or hemispherically, the past data are too poor or too sparse or both. Why not pick the best proxies and compute comparable modern temperatures to compare to the proxies at the specific proxy locations? It is easier to lessen resolution than to increase it.
Rosenthal, et al.’s temperature reconstruction, plotted in Figure 1, shows ~500-meter temperatures from the “Indonesian Throughflow” (Rosenthal, Linsley, & Oppo, 2013). Their data are from sediment cores taken in the Makassar Strait east of Borneo Island. This strait is a portion of the main connection between the Indian, Southern and Pacific Oceans. Its temperature is reflective of the temperature of significant portions of these three large water masses at 500 meters.
Rosenthal, et al. used an Mg/Ca (Magnesium/Calcium ratio) proxy from benthic foraminifers to estimate the 500-meter temperatures and they claim a temperature accuracy of ±0.7°C. Dating was done using radiocarbon (14C) and is probably no better than ±50 years. The radiocarbon dating was checked using distinctive volcanic ash layers and lead isotope dates (210Pb).
The fossils and shells studied by Rosenthal and his colleagues are from a bottom-dwelling foraminifer that lives at about 500 meters depth in the strait. Sea level varied over his study period, but he has corrected for this. Five hundred meters is deep enough to be insulated from short-term weather fluctuations on the chaotic surface, but shallow enough to reflect longer-term surface climatic fluctuations. In addition, from the University of Hamburg we have an accurate modern temperature for the period from about 2006-2016, at 500 meters, of about 7.7°C, this modern temperature is identified on the plot with a red box. Rosenthal’s reconstruction only goes back to 7100 BC, and has a resolution of 20 years to 30 BC, and 50 years before then.
Figure 1 represents ocean mixed layer and upper deep ocean temperatures in three major oceans. These oceans contain more heat capacity than the entire atmosphere. The reconstruction in Figure 1 illustrates the steady drop in surface temperatures since the Holocene Climatic Optimum, which ended about 6,000 years ago. The period of cooling after the Holocene Climatic Optimum is called the Neoglacial. Figure 1 suggests that the average temperature from 2006-2016 is quite normal, even cool, relative to the past 2,000 or 1,000 years, and before then. This temperature is only applicable to the Makassar Strait, but it is connected to three large oceans.
Several historical events are noted on the plot to show how civilization has been affected, at least in part, by the increasingly cooler temperatures. Historical events are important climatic indicators since they are accurately dated, and often indicate climate changes over large areas. The “Little Ice Age” or LIA, was a very cold and miserable time for humanity. It contributed to the Black Death Plague, the end of the Viking settlements in Greenland, and to the burning and persecution of witches and Jews in Europe, as they were often blamed for the cold weather (Behringer, 2010, pp. 98, 128). This is the “pre-industrial” weather the alarmists want us to return to, and without the benefits of fossil fuels. Welcome to Hell!
The Medieval Warm Period (MWP), The Roman Warm Period (RWP), and the Holocene Climatic Optimum are all warmer in the Makassar Strait than today. The Neoglacial cooling period, after the Climatic Optimum is well defined at this site.
Bo Christiansen and Fredrick Ljungqvist’s reconstruction of the past 2,000 years (Christiansen & Ljungqvist, 2012) is also important. As discussed in the previous post, they avoid spatial regression in their reconstruction to preserve as much climatic variability as possible. Their reconstruction is shown in Figure 2. It is only for the extra-tropical Northern Hemisphere. The red box, plotted at the year 2000, is the average extra-tropical Northern Hemisphere HadCRUT5 average temperature from 1970-2020. In this case it is meant to be compared to Christiansen and Ljungqvist’s 50-year smoothed reconstruction. The HadCRUT5 anomaly has been moved from the HadCRUT5 zero (1961-1990) to the 1880-1960 zero point used by Christiansen and Ljungqvist. “Extra-tropical” includes all HadCRUT5 5°x5° cells from 27.5°N to 87.5°N, ignoring null cells. The average HadCRUT5 cell temperature is area-weighted by latitude.
Just like in Figure 1, we did not try and expand the proxy record to more than it is meant to cover, we reduced HadCRUT5 to the area the proxy record covers. In this case the HadCRUT5 50-year average fits the proxy reconstruction well and is also roughly the same temperature as the MWP. The faint lines in Figure 2 are yearly proxy temperatures, they show much more variability than the 50-year smoothed curve but are not as meaningful as the 50-year average in terms of climate change.
In Figure 2 we show some of the warming and cooling events documented in Soon, et al.’s 2003 papers for the past two millennia (Soon & Baliunas, 2003) and (Soon, Baliunas, & Legates, 2003c). Soon and colleague’s make the important point that climate is a local thing, it does not vary uniformly across the globe. Large scale indicators of climate, such as global glacial advances and retreats, suggest that globally the Medieval Warm Period (MWP) was warmer, and a time of global glacial retreat. The Little Ice Age (LIA) was a period of glacial advance and colder weather. But, looking at detailed records suggests that both warming and cooling existed during both periods, depending upon the location. Most of the world warmed during the MWP, but it cooled significantly around Taylor Dome in Antarctica. The world generally cooled in the LIA, but Switzerland and Antarctica warmed significantly in the decades around 1540AD and 1800AD, respectively. The peaks of these global climate anomalies were reached at different times in different places.
Comparing these reconstructions to Vinther’s 2009 Greenland reconstruction (see the first post, Figure 2, where it is compared to Antarctic temperatures), as we do in Figure 3, illustrates how climate varies by region and by hemisphere. These local variations confound the “global warming” narrative.
Vinther’s reconstruction is built using the average of the Agassiz and Renland ice cores in Greenland, after correcting them for elevation changes. The Agassiz core is not actually from Greenland, but on a neighboring island. Vinther’s paper has a map of the two sites. The central graph in Figure 3 is the Vinther reconstruction in actual degrees C. The red box is the Greenland temperature average from the populated HadCRUT5 5×5 degree cells closest to the Agassiz and Renland sites for 2000-2020. As you can see, it is not very anomalous, relative to the Vinther record. It is lower than the peaks in the MWP, RWP and one to two degrees lower than the Holocene Climatic Optimum.
There are a large number of historical references in the timeline shown in Figure 5 and we will not explain all of them here, they are well documented in earlier posts here and here. We will just make the point that significant local climate changes, the only climate changes that matter to people, are historical events that are often described in detail by the historians of the time and dated precisely. These historical descriptions can be more valuable than biological or ice core proxies. There are three significant Northern Hemisphere or global climate changes that deserve special mention.
Roman Warm Period
The Roman Warm Period (RWP) was a period from roughly 100 BC to 200 AD, depending upon where you are. This was the time when robust civilizations developed in the Americas, around the Mediterranean, China, and India. Ch’in unified China by 200BC and Alexander invaded India just 136 years earlier. The Mayan civilization rose to prominence before 250 AD in present-day Mexico, Guatemala, and Belize. The Roman Warm Period truly marks the beginning of modern civilization, written records document all major events over most of the world since this time. These writings and most recent reconstructions suggest that temperatures, at least in the Northern Hemisphere, were warmer than today.
Medieval Warm Period
The Medieval Warm Period (MWP) is normally given as 800 AD to 1250 AD, but it began and ended at different times in different places. In the beginning of this period, temperatures in Central Greenland rose by about 1.5°C in about 200 years, but they are erratic. It has been well documented as a worldwide event, but is not synchronous. It is uncertain what the global average temperature was during the period and whether the world was warmer then, than now. But, certainly in many areas where we have records, such as Greenland, the UK, and China, temperatures were comparable to today and in many cases warmer. During this period, the Vikings were a dominant force in Europe and in the Middle East.
Little Ice Age
The Little Ice Age (LIA) was not a true ice age, but a cooler period that followed the Medieval Warm Period. It is generally considered to have started by 1350 AD and it ended between 1850 and 1900 AD. The LIA was the coldest period in the Holocene. Northern Hemisphere temperatures dropped from 1°C to 1.5°C on average, but, like the MWP, it was not synchronous around the world. Many areas in the Southern Hemisphere were warmer than normal, especially in Antarctica. It was not cold over the entire period, but the Little Ice Age saw many periods that were very cold, from the famous year without a summer (1816), to the great famine of 1315. New York Harbor completely froze over in 1780, the Norse colonies in Greenland starved and had to be abandoned. A recent study notes several droughts in Europe during the Little Ice Age. These occurred in AD 1540, 1590, 1626 and 1719, plus an especially intense drought from 1437-1473 AD.
Modern Warm Period
The Modern Warm Period starts between 1850 and 1905, which is also the time when people began systematically recording and collecting objectively calibrated instrumental surface air temperature data from around the world. These temperatures were spotty in the beginning, but by the middle of the 20th Century a good, land-based worldwide temperature database was developing. In 1979, satellites were launched that could give us a reasonably accurate and complete lower troposphere temperature record over nearly the entire globe. A discussion of the accuracy of the satellite temperature measurements can be found in an interesting paper by John Christy, Roy Spencer, and William Braswell (Christy, Spencer, & Braswell, 2000) here. Satellite data suggests the lower troposphere is warming at an unimpressive rate of 0.14°C/decade.
Modern global instrumental temperatures have only been available for a short time. Even if the entire 170-year record is used, it is too short to be representative of documented temperature extremes seen over the past 2,000 years. New York Harbor has not frozen over recently, and many areas now covered in glacial ice were ice-free in the MWP.
Going further back, to the beginning of the Holocene, temperature proxies are very sparse and constructing a hemispheric or global temperature reconstruction is futile. The proxies are not accurate that far back, and there is no way to establish that the proxy to temperature functions used will work that far back in time with only a 170-year calibration period.
As we show in this report and as recommended by Soon, et al. (Soon, Baliunas, Idso, Idso, & Legates, 2003b), it is far better to deal with proxies one at a time. Combining them statistically is misleading. When the exact location of the proxy record is known, modern temperature data is accurate and dense enough to extract a reasonable modern instrumental temperature record for the location at a matching temporal resolution. In the examples shown in this post, modern temperatures appear to be well within the range of natural variability for the past 2,000 years and the past 15,000 years.
It is well known that insolation varies by latitude and the records discussed in this post support that. Carbon dioxide is a well-mixed gas and one would expect it to affect global average temperature change approximately evenly over some sufficient time period. We see no evidence that this is the case now, but the record is short.
In the words of Professor Steven Mithen (Mithen, 2003, p. 507):
“The next century of human-made global warming is predicted to be far less extreme than that which occurred at [9600 BC]. At the end of the Younger Dryas, mean global temperature had risen by 7°C in fifty years, whereas the predicted rise for the next hundred years is less than 3°C. The end of the last ice age led to a 120 meter increase in sea level, whereas that predicted for the next fifty years is a paltry 32 centimeters at most…”
Perspective is important, we must recognize that the climate and temperature change we have observed over the past century are very tiny relative to past natural changes. To gain that perspective we must do valid comparisons of historical climate changes to today. That means local comparisons, not global. That means recognizing the poor temporal resolution of proxies and their questionable accuracy, which degrades with time. It is also important to recognize that while the Northern Hemisphere network of land-based weather stations has been good for some time, a good network in the Southern Hemisphere is very new. Finally, decent temperature records of the oceans are a very recent addition. Since oceans cover 70% of Earth’s surface, they are a reliable global temperature record, a record that has been underutilized to date. To put modern temperatures into historical perspective, we should go local, not global.
Download the bibliography here.
11 thoughts on “How to compare today to the past”
Dear Mr. May,
I have a question about your Figures 1 and 2. Could you please reply to firstname.lastname@example.org ?
Why not just ask it here?