A critical evaluation of volcanic cooling
Rau, Greg
www.atmos-chem-phys.net/13/3997/2013/
doi:10.5194/acp-13-3997-2013
© Author(s) 2013. This work is distributed
under the Creative Commons Attribution 3.0 License.
An empirical model of global climate – Part 1: A critical evaluation of volcanic cooling
T. Canty1, N. R. Mascioli1,*, M. D. Smarte2,**, and R. J. Salawitch1,2,3
1Department of Atmospheric and Oceanic Science, University of Maryland, College Park, MD, USA
2Department of Chemistry and Biochemistry, University of Maryland, College Park, MD, USA
3Earth System Science Interdisciplinary Center, University of Maryland, College Park, MD, USA
*now at: Department of Earth and Environmental Sciences, Columbia University, New York, NY, USA
**now at: Division of Chemistry and Chemical Engineering, California Institute of Technology, Pasadena, CA, USA
Abstract. Observed reductions in Earth's surface temperature following explosive volcanic eruptions have been used as a proxy for geoengineering of climate by the artificial enhancement of stratospheric sulfate. Earth cools following major eruptions due to an increase in the reflection of sunlight caused by a dramatic enhancement of the stratospheric sulfate aerosol burden. Significant global cooling has been observed following the four major eruptions since 1900: Santa María, Mount Agung, El Chichón and Mt. Pinatubo, leading IPCC (2007) to state "major volcanic eruptions can, thus, cause a drop in global mean surface temperature of about half a degree Celsius that can last for months and even years". We use a multiple linear regression model applied to the global surface temperature anomaly to suggest that exchange of heat between the atmosphere and ocean, driven by variations in the strength of the Atlantic Meridional Overturning Circulation (AMOC), has been a factor in the decline of global temperature following these eruptions. The veracity of this suggestion depends on whether sea surface temperature (SST) in the North Atlantic, sometimes called the Atlantic Multidecadal Oscillation, but here referred to as Atlantic Multidecadal Variability (AMV), truly represents a proxy for the strength of the AMOC. Also, precise quantification of global cooling due to volcanoes depends on how the AMV index is detrended. If the AMV index is detrended using anthropogenic radiative forcing of climate, we find that surface cooling attributed to Mt. Pinatubo, using the Hadley Centre/University of East Anglia surface temperature record, maximises at 0.14 °C globally and 0.32 °C over land. These values are about a factor of 2 less than found when the AMV index is neglected in the model and quite a bit lower than the canonical 0.5 °C cooling usually attributed to Pinatubo. This result is driven by the high amplitude, low frequency component of the AMV index, demonstrating that reduced impact of volcanic cooling upon consideration of the AMV index is driven by variations in North Atlantic SST that occur over time periods much longer than those commonly associated with major volcanic eruptions. The satellite record of atmospheric temperature from 1978 to present and other century-long surface temperature records are also consistent with the suggestion that volcanic cooling may have been over estimated by about a factor of 2 due to prior neglect of ocean circulation. Our study suggests a recalibration may be needed for the proper use of Mt. Pinatubo as a proxy for geoengineering of climate. Finally, we highlight possible shortcomings in simulations of volcanic cooling by general circulation models, which are also being used to assess the impact of geoengineering of climate via stratospheric sulfate injection.
Final Revised Paper (PDF, 9366 KB) Supplement (828 KB) Discussion Paper (ACPD)
Citation: Canty, T., Mascioli, N. R., Smarte, M. D., and Salawitch, R. J.: An empirical model of global climate – Part 1: A critical evaluation of volcanic cooling, Atmos. Chem. Phys., 13, 3997-4031, doi:10.5194/acp-13-3997-2013, 2013. Bibtex EndNote Reference Manager XML
Alan Robock
Robock, Alan, 2005: Comment on “Climate forcing by the volcanic eruption of Mount Pinatubo” by David H. Douglass and Robert S. Knox. Geophys. Res. Lett., 32, L20711, doi:10.1029/2005GL023287. http://climate.envsci.rutgers.edu/pdf/DouglassKnoxComment2005GL023287.pdf
Wigley, T. M. I., C. M. Ammann, B. D. Santer, and K. E. Taylor (2005a), Comment on ‘‘Climate forcing by the volcanic eruption of Mount Pinatubo’’ by David H. Douglass and Robert S. Knox, Geophys. Res. Lett., 32, L20709, doi:10.1029/2005GL023312.
Alan Robock Alan Robock, Professor II Editor, Reviews of Geophysics Director, Meteorology Undergraduate Program Associate Director, Center for Environmental Prediction Department of Environmental Sciences Phone: +1-848-932-5751 Rutgers University Fax: +1-732-932-8644 14 College Farm Road E-mail: rob...@envsci.rutgers.edu New Brunswick, NJ 08901-8551 USA http://envsci.rutgers.edu/~robock http://twitter.com/AlanRobock
--
You received this message because you are subscribed to the Google Groups "geoengineering" group.
To unsubscribe from this group and stop receiving emails from it, send an email to geoengineerin...@googlegroups.com.
To post to this group, send email to geoengi...@googlegroups.com.
Visit this group at http://groups.google.com/group/geoengineering?hl=en.
For more options, visit https://groups.google.com/groups/opt_out.
Alan Robock
Dear Greg,
I agree with Alan - Canty et al. must have a really serious error in what they've done. There is no way that the Pinatubo-induced cooling signal is only 0.14 degrees C.
The abstract for the paper makes absolutely no sense to me.
With best regards,
Ben
Alan Robock Alan Robock, Distinguished Professor Editor, Reviews of Geophysics Director, Meteorology Undergraduate Program Associate Director, Center for Environmental Prediction Department of Environmental Sciences Phone: +1-848-932-5751
Rutgers University Fax: +1-732-932-8644 14 College Farm Road E-mail: rob...@envsci.rutgers.edu New Brunswick, NJ 08901-8551 USA http://envsci.rutgers.edu/~robock http://twitter.com/AlanRobock
Much thanks to the colleague who informed us of Alan Robock’s comment on our paper.
We strongly encourage anyone interested in geo-engineering of climate to read this paper, despite Alan’s proclamation the paper is “fundamentally wrong” because we “double count the impact of volcanic eruptions on climate”.
In Canty et al. we show that if temporal variations in North Atlantic SST truly represent a proxy for variations in the strength of the Atlantic Meridional Overturning Circulation (AMOC), then the climate record from 1900 to present can be fit remarkably well by a model that represents increases in radiative forcing (RF) due to rising levels of GHGs, decreases in RF due to both tropospheric aerosols (pollution) and stratospheric aerosols (volcanoes), changes in RF caused by the ~11 year cycle in total solar irradiance, as well as variations in the exchange of heat between the atmosphere and ocean due to ENSO, the AMOC, and the long-term rise in oceanic temperature.
One consequence of fitting the climate record with a proxy for variations in the strength of the AMOC is a factor of 2 reduction in the cooling attributed to major volcanic eruptions.
Alan is critical of our work because he apparently believes North Atlantic SSTs dropped in direct response to major volcanic eruptions, such as Mt. Pinatubo. If this were true, our use of North Atlantic SST as a proxy for variations in the strength of the AMOC would be flawed.
Fourier analysis (Section 4.1.2 of Canty et al.) and conditional regression calculations (Section 4.1.3) were added to the final version of our paper to address Alan’s concern. We show that the diminution of volcanic cooling upon introduction of an index representing variations in the strength of the AMOC is driven by the low frequency, high amplitude component of North Atlantic SST. As discussed in our paper, this low frequency, high amplitude component of this climate forcing, sometimes referred to as the Atlantic Multidecadal Variability (AMV) and more often called the Atlantic Multidecdal Oscillation, is likely driven by variations in the density of seawater in the deepwater formation regions of the North Atlantic. In the comment Alan submitted to our discussion paper, he wrote “there is no proof that there is a physical mechanism with a 50-70 year period, because the data record is not long enough”, which apparently was motivated by his erroneous belief that the purely periodic behavior of the AMV was somehow important to our study. Conditional regression (CR) analysis reinforces the central premise of our paper: this technique yields a best estimate of 0.18°C for the maximum drop in global mean surface temperature attributed to the eruption of Mt. Pinatubo, with 95% confidence limits lying between 0.05 and 0.34°C. CR is an important complement to multiple linear regression because CR allows quantification of possible errors due to co-linearity of regressor variables.
Alan stated our paper “does not prove that the impact of volcanic eruptions is much smaller than previously thought nor that climate sensitivity is much smaller than commonly accepted”. On this we agree: our paper certainly does not prove either point!
“Prove” is a strong word that no reasonable scientist would ever use in association with any single study!!
Our study does, however, suggest the impact of volcanic eruptions on global climate is considerably smaller than commonly thought. Page 97 of the IPCC (2007) Physical Science Basis document states “Major volcanic eruptions can thus cause a drop in mean global surface temperature of about half a degree Celsius that can last for months or even years”. This notion is reinforced by Crutzen (Climate Change, 2006), who wrote the eruption of Mt Pinatubo led to “enhanced reflection of solar radiation to space by particles” that “cooled the earth’s surface on average by 0.5°C in the year following the eruption.”
Alan wrote “I recommend that you just ignore” our paper. We’d like to point out the following other papers that also contradict the statements quoted above regarding the strength of volcanic cooling. Figure 2 of Lean and Rind (GRL, 2008), Figure 8 of Thompson et al. (J. Climate, 2009), and Figure 7 (blue curve, middle panel) of Foster and Rahmstorf (ERL, 2011) all suggest the eruption of Mt. Pinatubo caused global average surface temperature to drop by 0.3°C, which is exactly what we find when variations in the strength of the AMOC are not considered in our model framework. The difference between our study and these other three papers is we focus on quantification of volcanic cooling, whereas the other studies focused on removing the effect on global temperature of RF due to volcanoes and other naturally varying components of the climate system (i.e., ENSO, solar cycle) so that the warming due to rising levels of GHGs can be precisely quantified. We have shown that, upon consideration of variations in the strength of the AMOC as another driver of climate, the best estimate for the cooling attributed to Pinatubo falls to ~0.15°C. As documented in Canty et al., North Atlantic SST began to fall well before the eruptions of Santa María, El Chichón, and Mt. Pinatubo.
Alan also criticizes our paper for not having proved “climate sensitivity is much smaller than commonly accepted”. We state (Section 5.1) that equilibrium climate sensitivity (ECS) is 1.9°C for our baseline simulation that relies on the data record of Church et al. (GRL, 2011) to define ocean heat export (OHE) and rises to 2.4°C if OHE is based instead on the observational record of Gouretski and Reseghetti (Deep-Sea Res., 2010). These estimates of ECS are closely aligned with another recent empirical estimate: Otto et al. (Nature Geoscience, 2013) state “the most likely value of ECS based on the energy budget of the most recent decade is 2.0°C, with a 5–95% confidence interval of 1.2–3.9 °C, compared with the 1970–2009 estimate of 1.9 °C (0.9–5.0 °C)”. The empirical best estimates of ECS from our study as well as Otto et al. (2013) are considerably less than found within most climate models (e.g., Figure 22 of Shindell et al., ACP, 2013). As quantified in our companion paper (Mascioli et al. doi:10.5194/acpd-12-23913-2012), the overestimate of ECS within climate models is likely due to excess OHE that results in the need for large climate feedbacks to match the observed temperature record.
We conclude by highlighting two additional points discussed in much greater detail within our paper.
The canonical ~0.5°C cooling due to the eruption of My Pinatubo is based on analysis of temperature measurements in the lower atmosphere, not the surface, obtained by the Microwave Sounding Unit (MSU) (Soden et al., Science, 2002). As shown in Canty et al. and Figure 7 (middle panel) of Foster and Rahmstorf (ERL, 2011), the decrease in lower atmospheric temperature following Pinatubo exceeds the surface cooling due to the smaller heat capacity of the atmosphere. Yet IPCC (2007) and Crutzen (2006) both state the ~0.5°C cooling induced by Pinatubo applies to global mean surface temperature. This statement is not supported by the observational record. The MSU time series of Soden et al. (2002) extended to present day is too short to allow definitive separation of atmospheric cooling induced by volcanic eruptions from that induced by variations in the strength of the AMOC. Nonetheless, Figures 13 and 14 of Canty et al. illustrate a plausible simulation of the MSU temperature record that supports our central premise that a portion of the observed cooling following the eruption of Mt. Pinatubo was indeed induced by variations in the strength of the AMOC that had begun to be manifest prior to the eruption.
Calculations from several climate models indicate major volcanic eruptions should have caused a ~0.5°C reduction in global mean surface temperature. If our suggestion regarding the strength of volcanic cooling from the climate record is correct, then what might be wrong with these calculations? In Section 5 of Canty et al., we discuss Figure 9.14 of IPCC (2007) that shows most climate models collapse the tropopause after major volcanic eruptions. This response is not seen in the ERA-40 meteorological reanalysis. Possibly climate models erroneously place a significant fraction of the volcanic aerosol below the tropopause: many simulations are based on prescribed volcanic aerosol loading given as a function of altitude, latitude, and time without regard for the height of the tropopause. Perhaps the simulated response of stratospheric circulation to volcanic perturbation is deficient: most of the GCMs used to estimate volcanic cooling did not participate in a rigorous evaluation of stratospheric dynamics (Eyring, Shepherd and Waugh (eds.): SPARC Report on the Evaluation of Chemistry-Climate Models, SPARC Report No. 5, WCRP-132, WMO/TD-No. 1526, 2010). Figures 15 and S11 of Canty et al. show time series of top of the atmosphere radiative flux measurements obtained by the Earth Radiation Budget Experiment (ERBE) in the solar and thermal portions of the spectrum. The influence of Pinatubo is barely discernable poleward of 20° latitude in either hemisphere because increased trapping of thermal radiation by volcanic aerosol nearly balances enhanced reflection of solar radiation. Yet modeling studies such as Stenchikov et al. (JGR, 2009) apply a global 3 W/m2 reduction in solar radiation that “dominates all other forcings for at least two years”. We suggest the atmospheric science community’s understanding of geo-engineering would be advanced if the volcanically induced influence on the height of the tropopause within climate models, as well as volcanic perturbation to the net top of the atmosphere radiative budget, were to be critically evaluated using observations within projects such as GeoMIP.
If members of the Geo-Engineering Google Group would like to read more, please have a look at our paper, available at: http://www.atmos-chem-phys.net/13/3997/2013/acp-13-3997-2013.html
Ross Salawitch and Tim Canty
Citation: Canty, T., Mascioli, N. R., Smarte, M. D., and Salawitch, R. J.: An empirical model of global climate – Part 1: A critical evaluation of volcanic cooling, Atmos. Chem. Phys., 13, 3997-4031, doi:10.5194/acp-13-3997-2013, 2013. Bibtex EndNote Reference Manager XML
--
Fred Zimmerman
Much thanks to the colleague who informed us of Alan Robock’s comment on our paper.
We strongly encourage anyone interested in geo-engineering of climate to read this paper, despite Alan’s proclamation the paper is “fundamentally wrong” because we “double count the impact of volcanic eruptions on climate”.
In Canty et al. we show that if temporal variations in North Atlantic SST truly represent a proxy for variations in the strength of the Atlantic Meridional Overturning Circulation (AMOC), then the climate record from 1900 to present can be fit remarkably well by a model that represents increases in radiative forcing (RF) due to rising levels of GHGs, decreases in RF due to both tropospheric aerosols (pollution) and stratospheric aerosols (volcanoes), changes in RF caused by the ~11 year cycle in total solar irradiance, as well as variations in the exchange of heat between the atmosphere and ocean due to ENSO, the AMOC, and the long-term rise in oceanic temperature.
One consequence of fitting the climate record with a proxy for variations in the strength of the AMOC is a factor of 2 reduction in the cooling attributed to major volcanic eruptions.
Alan is critical of our work because he apparently believes North Atlantic SSTs dropped in direct response to major volcanic eruptions, such as Mt. Pinatubo. If this were true, our use of North Atlantic SST as a proxy for variations in the strength of the AMOC would be flawed.
Fourier analysis (Section 4.1.2 of Canty et al.) and conditional regression calculations (Section 4.1.3) were added to the final version of our paper to address Alan’s concern. We show that the diminution of volcanic cooling upon introduction of an index representing variations in the strength of the AMOC is driven by the low frequency, high amplitude component of North Atlantic SST. As discussed in our paper, this low frequency, high amplitude component of this climate forcing, sometimes referred to as the Atlantic Multidecadal Variability (AMV) and more often called the Atlantic Multidecdal Oscillation, is likely driven by variations in the density of seawater in the deepwater formation regions of the North Atlantic. In the comment Alan submitted to our discussion paper, he wrote “there is no proof that there is a physical mechanism with a 50-70 year period, because the data record is not long enough”, which apparently was motivated by his erroneous belief that the purely periodic behavior of the AMV was somehow important to our study. Conditional regression (CR) analysis reinforces the central premise of our paper: this technique yields a best estimate of 0.18°C for the maximum drop in global mean surface temperature attributed to the eruption of Mt. Pinatubo, with 95% confidence limits lying between 0.05 and 0.34°C. CR is an important complement to multiple linear regression because CR allows quantification of possible errors due to co-linearity of regressor variables.
Alan stated our paper “does not prove that the impact of volcanic eruptions is much smaller than previously thought nor that climate sensitivity is much smaller than commonly accepted”. On this we agree: our paper certainly does not prove either point!
“Prove” is a strong word that no reasonable scientist would ever use in association with any single study!!
Our study does, however, suggest the impact of volcanic eruptions on global climate is considerably smaller than commonly thought. Page 97 of the IPCC (2007) Physical Science Basis document states “Major volcanic eruptions can thus cause a drop in mean global surface temperature of about half a degree Celsius that can last for months or even years”. This notion is reinforced by Crutzen (Climate Change, 2006), who wrote the eruption of Mt Pinatubo led to “enhanced reflection of solar radiation to space by particles” that “cooled the earth’s surface on average by 0.5°C in the year following the eruption.”
Alan wrote “I recommend that you just ignore” our paper. We’d like to point out the following other papers that also contradict the statements quoted above regarding the strength of volcanic cooling. Figure 2 of Lean and Rind (GRL, 2008), Figure 8 of Thompson et al. (J. Climate, 2009), and Figure 7 (blue curve, middle panel) of Foster and Rahmstorf (ERL, 2011) all suggest the eruption of Mt. Pinatubo caused global average surface temperature to drop by 0.3°C, which is exactly what we find when variations in the strength of the AMOC are not considered in our model framework. The difference between our study and these other three papers is we focus on quantification of volcanic cooling, whereas the other studies focused on removing the effect on global temperature of RF due to volcanoes and other naturally varying components of the climate system (i.e., ENSO, solar cycle) so that the warming due to rising levels of GHGs can be precisely quantified. We have shown that, upon consideration of variations in the strength of the AMOC as another driver of climate, the best estimate for the cooling attributed to Pinatubo falls to ~0.15°C. As documented in Canty et al., North Atlantic SST began to fall well before the eruptions of Santa María, El Chichón, and Mt. Pinatubo.
Alan also criticizes our paper for not having proved “climate sensitivity is much smaller than commonly accepted”. We state (Section 5.1) that equilibrium climate sensitivity (ECS) is 1.9°C for our baseline simulation that relies on the data record of Church et al. (GRL, 2011) to define ocean heat export (OHE) and rises to 2.4°C if OHE is based instead on the observational record of Gouretski and Reseghetti (Deep-Sea Res., 2010). These estimates of ECS are closely aligned with another recent empirical estimate: Otto et al. (Nature Geoscience, 2013) state “the most likely value of ECS based on the energy budget of the most recent decade is 2.0°C, with a 5–95% confidence interval of 1.2–3.9 °C, compared with the 1970–2009 estimate of 1.9 °C (0.9–5.0 °C)”. The empirical best estimates of ECS from our study as well as Otto et al. (2013) are considerably less than found within most climate models (e.g., Figure 22 of Shindell et al., ACP, 2013). As quantified in our companion paper (Mascioli et al. doi:10.5194/acpd-12-23913-2012), the overestimate of ECS within climate models is likely due to excess OHE that results in the need for large climate feedbacks to match the observed temperature record.
We conclude by highlighting two additional points discussed in much greater detail within our paper.
The canonical ~0.5°C cooling due to the eruption of My Pinatubo is based on analysis of temperature measurements in the lower atmosphere, not the surface, obtained by the Microwave Sounding Unit (MSU) (Soden et al., Science, 2002). As shown in Canty et al. and Figure 7 (middle panel) of Foster and Rahmstorf (ERL, 2011), the decrease in lower atmospheric temperature following Pinatubo exceeds the surface cooling due to the smaller heat capacity of the atmosphere. Yet IPCC (2007) and Crutzen (2006) both state the ~0.5°C cooling induced by Pinatubo applies to global mean surface temperature. This statement is not supported by the observational record. The MSU time series of Soden et al. (2002) extended to present day is too short to allow definitive separation of atmospheric cooling induced by volcanic eruptions from that induced by variations in the strength of the AMOC. Nonetheless, Figures 13 and 14 of Canty et al. illustrate a plausible simulation of the MSU temperature record that supports our central premise that a portion of the observed cooling following the eruption of Mt. Pinatubo was indeed induced by variations in the strength of the AMOC that had begun to be manifest prior to the eruption.
Calculations from several climate models indicate major volcanic eruptions should have caused a ~0.5°C reduction in global mean surface temperature. If our suggestion regarding the strength of volcanic cooling from the climate record is correct, then what might be wrong with these calculations? In Section 5 of Canty et al., we discuss Figure 9.14 of IPCC (2007) that shows most climate models collapse the tropopause after major volcanic eruptions. This response is not seen in the ERA-40 meteorological reanalysis. Possibly climate models erroneously place a significant fraction of the volcanic aerosol below the tropopause: many simulations are based on prescribed volcanic aerosol loading given as a function of altitude, latitude, and time without regard for the height of the tropopause. Perhaps the simulated response of stratospheric circulation to volcanic perturbation is deficient: most of the GCMs used to estimate volcanic cooling did not participate in a rigorous evaluation of stratospheric dynamics (Eyring, Shepherd and Waugh (eds.): SPARC Report on the Evaluation of Chemistry-Climate Models, SPARC Report No. 5, WCRP-132, WMO/TD-No. 1526, 2010). Figures 15 and S11 of Canty et al. show time series of top of the atmosphere radiative flux measurements obtained by the Earth Radiation Budget Experiment (ERBE) in the solar and thermal portions of the spectrum. The influence of Pinatubo is barely discernable poleward of 20° latitude in either hemisphere because increased trapping of thermal radiation by volcanic aerosol nearly balances enhanced reflection of solar radiation. Yet modeling studies such as Stenchikov et al. (JGR, 2009) apply a global 3 W/m2 reduction in solar radiation that “dominates all other forcings for at least two years”. We suggest the atmospheric science community’s understanding of geo-engineering would be advanced if the volcanically induced influence on the height of the tropopause within climate models, as well as volcanic perturbation to the net top of the atmosphere radiative budget, were to be critically evaluated using observations within projects such as GeoMIP.
If members of the Geo-Engineering Google Group would like to read more, please have a look at our paper, available at: http://www.atmos-chem-phys.net/13/3997/2013/acp-13-3997-2013.html
Ross Salawitch and Tim Canty
Citation: Canty, T., Mascioli, N. R., Smarte, M. D., and Salawitch, R. J.: An empirical model of global climate – Part 1: A critical evaluation of volcanic cooling, Atmos. Chem. Phys., 13, 3997-4031, doi:10.5194/acp-13-3997-2013, 2013. Bibtex EndNote Reference Manager XML--