Annotated bibliography for "The Great Tambora Eruption in 1815 and Its Aftermath" written by Richard B. Stothers

by

Kyle Nichols

Angell, J.K. and J. Korshover. 1985. Surface Temperature Changes Following the Six Major Volcanic Episodes between 1780 and 1980. Journal of Climate and Applied Meteorology, 24, 937-951.

This article uses a statistical analysis of worldwide temperatures following the six eruptions with large dust-veil index. The temperature recording locations were clustered in Europe with New Haven, Connecticut the only station in North America. Each station has a record extending to at least 1781. The authors use yearly temperatures, because they state that large volcanic eruptions should affect temperatures for at least 1-2 years. Temperature depressions after the 1783 Asama-Laki eruptions are evident for up to three years after the eruptions. Evidence for temperature depressions following the Tambora eruption is only present in the New Haven data. The European data only shows a one year depression followed by a rise in temperatures. The authors conclude that Tambora did not have hemisphere wide temperature depressions. It is noted that a review of the paper suggested that temperature variability is most affected by volcanic eruptions and not the mean annual temperature. This study does not use this method. Conseguina, Nicaragua eruption of 1835 indicates temperature depressions greatest in New Haven data. European data are less suggestive, and a long cooling trend in the hemisphere data indicates that the temperature decreases are not necessarily related to the Conseguina eruption. The authors do state that these data are the second most impressive in the study, but I do not see it. All sites show an increase in temperature following the eruption of Krakatau in 1883, yet the hemisphere data shows a decrease. One would have to look 2-4 years after the eruption to observe a decrease in mean temperatures. Other methods by other researchers do show decreases in hemisphere temperatures. The data for temperature depressions following the series of the 1902 eruptions of Santa Maria, Soufriere and Pelee are inconclusive for all temperature recording sites. The Agung eruption of 1963 suggests hemisphere wide temperature depressions but evidence at the temperature recording sites is inconclusive.

The statistical analysis indicates that out of a sample population of 96, there are only 27 significant (5% confidence level) temperature depression occurrences. There were no significant temperature elevations. This indicates that large eruptions do not guarantee temperature decreases. El Nino events can dampen temperature response to large eruptions.

This paper uses many different data sources and different statistical methods. The data from the early 1800s is scarce, especially in North America. New Haven, CT. does not represent the whole continent, and thus I think that the findings of this paper are not robust. I think that other methods rather than just annual mean temperatures should examine effects of volcanic eruptions on temperature.

Dai, Jihong, Mosley-Thompson, Ellen and Lonnie G. Thompson. 1991. Ice Core Evidence for an Explosive Tropical Volcanic Eruption 6 Years Preceding Tambora. Journal of Geophysical Research, 96, 17,361-17,366.

The authors base the paper on finding evidence of elevated SO42- signatures in ice cores and relating these elevations to explosive volcanic eruptions, due to a lack of other known SO42- emplacement mechanisms. Two cores from Greenland and two cores from Antarctica show increases of SO42- concentrations for the well know Tambora eruption of 1815 and also for a period that is six years prior (1809) to the Tambora event. Both of these events demonstrate elevated SO42- concentrations lasting for only two years. The authors conclude that since the latter "event" (the 1809 event) is present in both hemispheres, the explosive eruption must have been within 20 of the equator, since only large tropical eruptions produce evidence that are present in ice core records of both Greenland and Antarctica. Furthermore, SO42- concentrations of the 1809 event are 40% - 60% of the Tambora SO42- concentrations. Although the authors state that calculating volcanic magnitudes based solely on ice cores is risky due to differences in deposition amounts and post-depositional surface processes, they conclude that this unknown eruption produced one fifth of the SO2 of Tambora and 2-3 times the amount of SO2 Krakatau.

I have reservations of the findings of this paper. All previous evidence, outside of the ice core world, does not support an eruption larger than Krakatau six years prior to Tambora. I think that the authors should have investigated, or at least mentioned the possibility of two smaller simultaneous eruptions occurring the same year in different parts of the globe (i.e. one in the northern hemisphere and one in the southern hemisphere). I also think that the estimation of the magnitude of the eruption is a leap of faith just using ice cores. Most importantly, there is no written record of a large eruption during this time period, and the authors do not have a possible (physical) volcano to account for the record. Some smaller eruptions can emplace larger amouts of SO2 without violently erupting large volumes of ash.

Hughes, Patrick. 1979. The Year without a Summer. Weatherwise, ??, 108-111.

Hughes introduces 1816 as "the year without a summer", "poverty year" and "eighteen hundred and froze-to-death." Frosts in June, July and August killed crops. The northeast was the hardest hit by the cold weather, mainly because the western frontier did not reach the Mississippi River yet and the weather recordings are non-existent.

April and May were both cold months in the northeast. Planting was delayed until after the frost left in early June. Farmers finally planted their fields only to have frost and snow kill the crops on June 6 ­9. Again the crops were planted after the cold weather and again were killed by cold weather (at the end of the first week of July). Crops were planted a third time after the cold weather ceased. The crops were killed a third time on August 20. The latest frost eliminated any chance to have a growing season and many people in the northeast were talking about famine. The following months and year exhibited the greatest exodus in the northeast's history. The cold weather was not limited to the summer months. The rest of the year was also below normal for temperature.

William Humphreys, a Weather bureau scientist, correlated the eruptions of Soufriere on St. Vincent Island in 1812, Mayon in the Philippines in 1814, and Tambora in 1815, with the depression of world wide temperatures. Humphreys stated that volcanic dust is 30 times more effective in keeping the suns radiation out than keeping the Earth's in. Furthermore, it can take years for the finest dust to settle out.

The theory of the time to explain the depression of temperatures was the sunspot theory. The sunspot theory accounts for the decrease in sun's radiation emitted to the Earth. Humphreys discounted this theory with the volcanic dust theory. The article states although scientists have learned much more about volcanic eruptions, there is not much more to be added to the impact upon weather and climate.

This article is a good historical perspective on what happened during the summer of 1816. There is not much scientific evidence to support the theory that volcanic eruptions depress world wide temperatures.

Ludlum, D.M. 1985. 1816: "A year without a summer". in The Vermont Weather Book, Vermont Historical Society, Montpelier, Vermont.

Accounts of the summer of 1816 were recorded mainly at educational institutions (Middlebury, Williams and Dartmouth) and by newspapers. 1811-1820 was the coldest recorded decade of the 19th Century. 1835-1837 was also an abnormally cool period, but the remainder of the decade was warm. The winter of 1816 was mild. A snow storm hit Vermont on June 7. July was cool but not devastating. August was warm until the third week when a devastating frost eliminated all hopes of growing corn. For the most part the fall was mild, except for a snow storm in October. Many people left Vermont for the Genesse valley near Rochester, New York. An atmospheric model of a high pressure sitting over Hudson's Bay in Canada accounts for the cold weather.

This paper is an interesting account of the summer of 1816. Besides the temperature depressions in the summer months, there are two obvious relations to the other papers I have read. The cold period of 1835-1837 is at the same time as the eruption of Coseguina in Nicaragua, possibly accounting for three cool years out of an otherwise warm decade. Also, the mild winter preceding temperature depressions is consistent with the other papers I have read. Maybe someone should tell Mr. Ludlum about the volcanic theory.

Pyle, D.M. 1997. The global impact of the Minoan eruption of Santorini, Greece. Environmental Geology, 30 (1/2), 59-61.

The Minoan eruption of Santorini was at least a factor of two smaller than the eruption of Tambora. The sulfur equivalent of the major eruptions is the main cause of temperature depressions (for 1-3 years) rather than silica ash. The sulfur ejected by the Minoan eruption is actually 2-3 times less than the mass of sulfur ejected by Mt. Pinatubo. Although, the frequency of erupted volume similar to the Minoan eruption is only 8 per 1000 years, the frequency of sulfur emitting eruptions similar to the Minoan eruption is 18 per1000 years (based on ice core evidence).

This paper simplifies the occurrence of large eruptions and the quantity of sulfur emitting eruptions. These events are fairly rare (about 2 per century) and suggest ,to me, that it is unlikely that volcanoes play a direct role in long-term climate change.

Rampino, M.R. 1989. Distant Effects of the Tambora Eruption of April 1815 An Eyewitness Account. EOS, 70, 1559-1560.

This paper is a response to a questionnaire sent out by the Lieut. governor of Java. The witness stated that up to six days before the climatic eruption sounds similar to cannons or thunder were heard. On Tuesday the 11th of April a tremulous motion of the earth was felt. The sun was barley visible through the ash. The dark ash cloud persisted until the 18th. Temperatures were cooler under the ash cloud. No serious effects to animals were described at this location. The crops were more susceptible to the ash fall. The eruption was thought to be from the Mountain Kelut, as told by the aged inhabitants.

An interesting paper of an eyewitness account. Indications of how slow news traveled are present by the confusion of the minor eruptions with cannons and thunder, and the misidentification of the volcano.

Rampino, M.R., Self, S. and R.W. Fairbridge. 1979. Can Rapid Climatic Change Cause Volcanic Eruptions? Science, 206, 826-829.

Within the past 30,000 years, some glacial advances began 300-700 years before significant volcanic activity, while other glacial advances began up to 1000 years after the volcanic activity. Historic climate data suggests that temperature depression is on the same order of background variations (<1C), and the cooling period only lasts about 1-3 years. The other argument states that recent eruptions have been small to moderate in size, or occurred at the wrong time or place to have significant affects. The decade of the Tambora eruption coincides with the most pronounced low in the mean sunspot record of the last 250 years. This indicates a lower solar output. The 1810s was also a time of low magnetic intensity. These are some other possible explanations for climatic cooling. Plots of temperature variations and large volcanic eruptions indicate that cooling does occur after the eruption, but the cooling trend is usually present before the eruption also. Error in dating techniques should be remembered when trying to accurately determine a cause effect relationship. The authors state that although, many believe eruptions cause cooler climates, they believe that the opposite is also possible. Large events (like Tambora) only occur once every 1000 years at best, and no geological evidence has been found to indicate that these events have occurred on decade intervals to have an accumulative effect on depressing global temperatures to start an ice age. Many volcanic records from sea cores in the Gulf of Mexico post-date the initiation of sharp temperature decreases indicated by biostratigraphic studies of the same cores. Some climate triggering volcanism mechanisms are: Redistribution of water gives rise to hydroisostatic and glaciosostatic readjustments; asymmetric loading of the earth requires the globe to adjust its spin on the axis causing crustal adjustments that would happen along plate margins and faults thus triggering eruptions; and hydrostatic unloading of the crust favors the movement of basaltic magma.

The authors do a great job of speculating that climate may trigger volcanism. Many of the hypothesis are purely theory with no data to substantiate their claims. I think that this paper is interesting in that it proposes what previously seemed to me as unthinkable hypothesis of the relation of climate and volcanism.

Rampino, M.R., and S. Self. 1982. Historic Eruptions of Tambora (1815), Krakatau (1883), and Agung (1963), Their Stratospheric Aerosols, and Climatic Impact. Quaternary Research, 18, 127-143.

The Tambora eruption was dominated by ignimbrites. The temperature depressions averaged 0.8C form 1815 to 1816. The summer of 1816 was the coldest on record at New Haven, Conn. From 1780 to 1968. Other large eruptions of 1810-1820 may have contributed to the total stratospheric-aerosol load. Krakatau induced temperature depressions were about 0.3 to 0.4C. Temperatures in the Southern Hemisphere actually increased after the Agung eruption, while temperatures in the Northern Hemisphere decreased 0.3 to 0.4C. Overall volcanic activity usually can decrease global temperatures a few tenths of a degree Celsius for up to three years. Long-term climatic effects are probably a result of secondary feedback mechanisms. Temperature records of the past 200 years indicates that temperature decreases are a usually a result of the sulfur aerosols in the atmosphere rather than the silicic dust. Ice cores best record small volume but possibly sulfur-rich eruptions, but ice-cores tend to overemphasize eruptions in certain regions. Large plinian eruptions (silicic magma) are usually relatively poor in SO2, but Tambora is an exception. Large amounts of fine ash in large eruptions might act "to quickly scavenge some of the volitiles from the eruption." Mt. St. Helens had minimal climatic effect (high latitude, low SO2, and possibly efficient scavenging of aerosols by the ash. Ratios of Tambora:Krakatau:Agung for volume of ash are 150:20:1 and for sulfate 7.5:3:1. The time of year and latitude are critical in controlling the extent of aerosol dispersal.

This is the landmark paper that suggests that aerosols are the controlling factor in temperature depressions. I think that the authors could have better addressed background variations of temperature.

Rampino, M.R., and S. Self. 1993. Climate-Volcanism Feedback and the Toba Eruption of ~74,000 Years Ago. Quaternary Research, 40, 269-280.

The Toba eruption emitted about 800 km3 of dense rock equivalent. It is inferred that much of the ash cloud came from pyroclastic flows and not from a plinian type eruption. These clouds were driven by convection cells. The sulfur in the atmosphere (an order of magnitude greater than Tambora) would generate an aerosol optical depth peak (of 1.3) within 5 months and would have a lifetime of 2 years. Modeled estimates of Toba induced temperature depression using historical eruptions predicts 4 + 1 C. Right after the eruption temps at mid-latitudes could have dropped as much as 5 ­15 C for short periods of time. Ocean cooling as a result of the eruption could prolong the cool period and induce a feedback mechanism with the ice caps that could produce decade long temperature depressions. There was already significant snow cover at the time of the Toba eruption and a drop of up to 10C at the critical inland locations could have been enough to accelerate the feedback to create and hold more snow (increased albedo etc.). Other models suggest rapid expansion of ice sheets without volcanic influence.

Eruptions of Indonesian volcanoes deposit thick ash layers in the deep-sea record at ~400,000 year intervals. These records happen to coincide with the earth's eccentricity of 413,000 years and the last grouping of ash at ~2.4 my correlates with the onset of sever glaciation. This phenomenon is found elsewhere. There is evidence that glaciations can induce volcanic eruptions. One example is that small pressure differences (mass loading) can cause the movement of less dense magma to the surface. Is it the chicken or the egg?

This paper is a great look at a larger eruption than Tambora that occurred at a critical time in glacial chronology. Large eruptions at critical periods might be able to accelerate glacial activity.

Robock, A. and J. Mao. 1995. The Volcanic Signal in Surface Temperature Observations. Journal of Climate, 8, 1086-1103.

The eruptions of El Chichon (1982) and Pinatubo (1991) produced a cooling force larger than all anthropogenic greenhouse gases at the time. Sulfur aerosols scatter the incoming radiation, while the volcanic dust veil absorbs both longwave and shortwave radiation. Temperature depressions are better observed by taking the effects of El Nino events out of the temperature record and by dividing the globe into latitude bands. At high latitudes in the Northern Hemisphere there is a warming trend for the first winter. The temperature depressions last only two years; this is shorter than previously calculated. El Nino events usually only affect the global temps for one year, thus volcanic effects are visible for at least one year.

This paper is a noble effort on the part of the authors to separate volcanic effects from El Nino effects. Only recent volcanic eruptions can be investigated, which still leaves questions pertaining to the older eruptions such as Tambora. More work still has to be accomplished better understand the relationship between volcanoes and El Ninos.

Self, S. and M.R. Rampino. 1984. Volcanological study of the great Tambora eruption of 1815. Geology, 12, 659-663.

Tambora eruption of April 10-11, 1815, caused the death of more than 90,000 people. The volcano erupted an estimated range of 30km3 to 300km3 of tephra. Tambora is a shield volcano of alkalic mafic to undersaturated intermediate lavas. Only one pyroclastic flow predates the 1815 event, suggesting that the volcano may be in a transitional stage in magmatic composition and eruptive nature from effusive to explosive. The top 1400 m of the volcano was lost during the eruption. The climatic event is characterized by a one hour plinian phase followed by subsequent pyroclastic flows that reached the sea. Plinian ash fall and pyroclastic flows were concurrent for the first 4 hours of the eruption, then pyroclastic flows dominated the later stages of eruption. The pyroclastic flows are probably the cause for tsunamis reported during the eruption. (hmm) More than 500,000 km2 were covered by >1 cm of ash and more than 1 million km2 probably experienced ash fallout. Due to sparse fall out data, estimation of eruption volume is difficult. The averaged value of 150 km3 seems to be the accepted value for the ash fall out and the authors estimate an additional 25 km3 for ignimbrites. The estimated magma volume is 50 km3. The average output rate of the magma is estimated to be 6 x 105 m3 s-1. Only because of eye-witness accounts do we know the true magnitude of the Tambora eruption. Many of the ash fall deposits are no longer visible. It leaves one to wonder what other large explosive eruptions have occurred in the past unbeknown to us.

Self et al. detail the eruption history of Tambora. This is a great paper to read about the findings but not to digest the methods of data reduction and calculations.