Abstract
The dynamics of a linear two-box energy balance climate model is analyzed as a fast–slow system, where the atmosphere, land, and near-surface ocean taken together respond within few years to external forcing whereas the deep-ocean responds much more slowly. Solutions to this system are approximated by estimating the system’s time-constants using a first-order expansion of the system’s eigenvalue problem in a perturbation parameter, which is the ratio of heat capacities of upper and lower boxes. The solution naturally admits an interpretation in terms of a fast response that depends approximately on radiative forcing and a slow response depending on integrals of radiative forcing with respect to time. The slow response is inversely proportional to the “damping-timescale”, the timescale with which deep-ocean warming influences global warming. Applications of approximate solutions are discussed: conditions for a warming peak, effects of an individual pulse emission of carbon dioxide (CO\(_{\mathrm{2}}\)), and metrics for estimating and comparing contributions of different climate forcers to maximum global warming.
Similar content being viewed by others
Notes
This can be seen by writing Eq. (1) as \(c_{\mathrm{s}}\frac{{\text {d}}T_{\mathrm{s}}}{{\text {d}}t}=-B\left( T_{\mathrm{s}},T_{\mathrm{d}}\right) -H\left( T_{\mathrm{s}},T_{\mathrm{d}}\right) +F(t)\), where \(B\left( T_{\mathrm{s}},T_{\mathrm{d}}\right) =\beta T_{\mathrm{s}}+\left( \eta -1\right) \gamma \left( T_{\mathrm{s}}-T_{\mathrm{d}}\right)\) describes change in radiative flux at the top of the atmosphere (Held et al. 2010) and \(H\left( T_{\mathrm{s}},T_{\mathrm{d}}\right) =\gamma \left( T_{\mathrm{s}}-T_{\mathrm{d}}\right)\). As the system approaches equilibrium, change in radiative flux varies only with \(T_{\mathrm{s}}\) as \(\beta T_{\mathrm{s}}\); but in the beginning of the response where \(T_{\mathrm{d}}\approx 0\), the sensitivity is higher, i.e. \(\left( \beta +\left( \eta -1\right) \gamma \right) T_{\mathrm{s}}\). The magnitude of this difference depends on the value of \(\eta\), with generally \(\eta >1\) (Held et al. 2010; Winton et al. 2010).
This is based on heat capacity of water of 4180 J kg\(^{{-1 }}\)K\(^{{-1}}\), density of seawater of 1030 kg m\(^{{-3}}\), and approximating that the oceans cover 70 % of Earth’s surface, following Geoffroy et al. (2013a).
We use estimates of \({\text {BC}}_{{\text {PI}}}=32\times 10^{6}\) kg and \(\tau _{{\text {BC}}}=10\) days, based on results in Skeie et al. (2011a); see Seshadri (2015) for details on estimation of these quantities. Present emissions and mean atmospheric burden of BC (for 2014) are estimated as \(8.6\times 10^{9}\) kg and \(200\times 10^{6}\) kg respectively, based on Skeie et al. (2011b) and following Seshadri (2015).
For N\(_{\mathrm{2}}\)O, preindustrial and present concentrations are taken as 0.27 ppmv and 0.32 ppmv respectively (Myhre et al. 2013), present emissions are 16.3 Tg N\(_{\mathrm{2}}\)O–N per year (Davidson and Kanter 2014), and atmospheric lifetime is 114 years (Pierrehumbert 2014). For CH\(_{\mathrm{4}}\), preindustrial and present concentrations are taken as 0.80 ppmv and 1.80 ppmv respectively (Myhre et al. 2013), present emissions are 350 Tg per year (Myhre et al. 2013), and atmospheric lifetime is 12 years (Pierrehumbert 2014).
\(B_{T_{\mathrm{s}}}=\beta +\left( \eta -1\right) \gamma\), \(B_{T_{\mathrm{d}}}=-\left( \eta -1\right) \gamma\), \(H_{T_{\mathrm{s}}}=\gamma\), and \(H_{T_{\mathrm{d}}}=-\gamma\), and the tilde-variables are simply these aforementioned variables divided by \(c_{\mathrm{s}}\).
These differ from eigenvalues estimated by applying the small-\(\varepsilon\) approximation to the characteristic polynomial of A, as can be shown. The perturbation approach above is essential to correctly estimate eigenvalues, because it alone preserves continuity of the eigenvectors’ components in the zero-\(\varepsilon\) limit.
One might also view this as the condition for “adjusted forcing” \(F_{{\text {adj}}}(t)=\eta \gamma T_{\mathrm{d}}(t)+F(t)\) to peak and then decline [compare with Eq. (1)]. This can be verified by substituting the approximation for \(T_{\mathrm{d}}(t)\) and comparing results with the aforementioned discussion. This clarifies the origin of the effect that is approximately characterized by the damping-timescale. When F(t) begins to decrease the deep-ocean temperature \(T_{\mathrm{d}}(t)\) is still increasing, and the rate of decrease in radiative forcing must be large enough to compensate so that \(F_{{\text {adj}}}(t)\) can peak. The introduction of \(F_{{\text {adj}}}(t)\) decouples the upper box from the deep-ocean response, and peaking of the corresponding forcing \(F_{{\text {adj}}}(t)\) leads the fast subsystem forced by it, i.e. the EBM’s upper box temperature, to also peak.
We treat \(m_{Pulse}\) and \(\triangle {\text {CO}}_{2,{\text {Pulse}}}(t)\) as having the same units, for example mass can be described in concentration units or concentration of CO\(_{\mathrm{2}}\) can be described by corresponding mass in the atmosphere.
It can be shown that at the critical point \(t=t_{wp}\) where \(\dot{T}_{\mathrm{s}}\left( t_{wp}\right) =0\), maximum warming from the pulse follows the simplified equation \(\frac{c_{\mathrm{s}}T_{\mathrm{s}}\left( t_{wp}\right) }{\tau _{2}}\cong F\left( t_{wp}\right) +\frac{1}{\tau _{D}}F_{1}\left( t_{wp}\right)\).
References
Allen MR, Stocker TF (2014) Impact of delay in reducing carbon dioxide emissions. Nat Clim Change 4:23–26. doi:10.1038/nclimate2077
Allen MR, Frame DJ, Huntingford C, Jones CD, Lowe JA, Meinshausen M, Meinshausen N (2009) Warming caused by cumulative carbon emissions towards the trillionth tonne. Nature 458:1163–1166. doi:10.1038/nature08019
Berglund N, Gentz B (2006) Noise-induced phenomena in slow–fast dynamical systems: a sample paths approach. Springer, Berlin
Boden T, Marland G, Andres B (2011) Global CO\(_2\) emissions from fossil-fuel burning, cement manufacture, and gas flaring, pp 1751–2008
Bowerman NHA, Frame DJ, Huntingford C, Lowe JA, Smith SM, Allen MR (2013) The role of short-lived climate pollutants in meeting temperature goals. Nat Clim Change 3:1021–1024. doi:10.1038/nclimate2034
Byrne B, Goldblatt C (2014) Radiative forcing at high concentration of well-mixed greenhouse gases. Geophys Res Lett 41:152–160. doi:10.1002/2013GL058456
Caldeira K, Casting JF (1993) Insensitivity of global warming potentials to carbon dioxide emission scenarios. Nature 366:251–253. doi:10.1038/366251a0
CCAC (2014) Short-lived climate pollutants. http://www.unep.org/ccac/Short-LivedClimatePollutants/Definitions/tabid/130285/Default.aspx
Davidson EA, Kanter D (2014) Inventories and scenarios of nitrous oxide emissions. Environ Res Lett 9:1–12. doi:10.1088/1748-9326/9/10/105012
Forster P, Ramaswamy V, Artaxo P, Berntsen T, Betts R, Fahey D (2007) Climate Change 2007: the physical science basis. Contribution of working group I to the fourth assessment report of the intergovernmental panel on climate change, Chap. changes in atmospheric constituents and in radiative forcing. Cambridge University Press, Cambridge
Geoffroy O, Saint-Martin D (2014) Pattern decomposition of the transient climate response. Tellus A 66:1–7. doi:10.3402/tellusa.v66.23393
Geoffroy O, Saint-Martin D, Olivie DJL, Voldoire A, Bellon G, Tyteca S (2013a) Transient climate response in a two-layer energy-balance model. Part I: analytical solution and parameter calibration using CMIP5 AOGCM experiments. J Clim 26:1841–1857. doi:10.1175/JCLI-D-12-00195.1
Geoffroy O, Saint-Martin D, Bellon G, Voldoire A, Olivie DJL, Tyteca S (2013b) Transient climate response in a two-layer energy-balance model. Part II: representation of the efficacy of deep-ocean heat uptake and validation for CMIP5 AOGCMs. J Clim 26:1859–1876. doi:10.1175/JCLI-D-12-00196.1
Gillett NP, Arora VK, Zickfeld K, Marshall SJ, Merryfield WJ (2011) Ongoing climate change following a complete cessation of carbon dioxide emissions. Nat Geosci 4:83–87. doi:10.1038/ngeo1047
Gregory JM (2000) Vertical heat transports in the ocean and their effect on time-dependent climate change. Clim Dyn 16:501–515. doi:10.1007/s003820000059
Held IM, Winton M, Takahashi K, Delworth T, Zeng F, Vallis GK (2010) Probing the fast and slow components of global warming by returning abruptly to preindustrial forcing. J Clim 23:2418–2427. doi:10.1175/2009JCLI3466.1
Hirsch MW, Smale S (1974) Differential equations, dynamical systems, and linear algebra. Academic Press, New York
Holmes MH (2013) Introduction to perturbation methods. Springer, New York
Joos F, Roth R, Fuglestvedt JS, Peters G, Brovkin V, Eby M, Edwards N, Eleanor B (2013) Carbon dioxide and climate impulse response functions for the computation of greenhouse gas metrics: a multi-model analysis. Atmos Chem Phys 13:2793–2825. doi:10.5194/acp-13-2793-2013
Lashof DA, Ahuja DR (1990) Relative contributions of greenhouse gas emissions to global warming. Nature 344:529–531. doi:10.1038/344529a0
Li C, von Storch J-S, Marotzke J (2013) Deep-ocean heat uptake and equilibrium climate response. Clim Dyn 40:1071–1086. doi:10.1007/s00382-012-1350-z
Long S-M, Xie S-P, Zheng X-T, Liu Q (2014) Fast and slow responses to global warming: sea surface temperature and precipitation patterns. J Clim 27:285–299. doi:10.1175/JCLI-D-13-00297.1
Lu J, Zhao B (2012) The role of oceaning feedback in the climate response to doubling CO\(_2\). J Clim 25:7544–7563. doi:10.1175/JCLI-D-11-00712.1
Marshall DP, Zanna L (2014) A conceptual model of ocean heat uptake under climate change. J Clim 27:8444–8465. doi:10.1175/JCLI-D-13-00344.1
Marshall J, Scott JR, Armour KC, Campin J-M, Kelley M, Romanou A (2015) The ocean’s role in the transient response of climate to abrupt greenhouse gas forcing. Clim Dyn 44:2287–2299. doi:10.1007/s00382-014-2308-0
Matthews HD, Gillett NP, Stott PA, Zickfeld K (2009) The proportionality of global warming to cumulative carbon emissions. Nature 459:829–832. doi:10.1038/nature08047
Meinshausen M, Meinshausen N, Hare W, Raper SCB, Frieler K, Knutti R, Frame DJ, Allen MR (2009) Greenhouse-gas emission targets for limiting global warming to 2C. Nature 458:1158–1162. doi:10.1038/nature08017
Meinshausen M, Smith SJ, Calvin K, Daniel JS, Kainuma MLT (2011) The RCP greenhouse gas concentrations and their extensions from 1765 to 2300. Clim Change 109:213–241. doi:10.1007/s10584-011-0156-z
Myhre G, Shindell D, Bréon F-M, Collins W, Fuglestvedt J, Huang J (2013) Climate change 2013: the physical science basis. In: Contribution of working group I to the fifth assessment report of the intergovernmental panel on climate change, Chap. anthropogenic and natural radiative forcing. Cambridge University Press, Cambridge, pp 659–740
Parry M, Lowe J, Hanson C (2009) Overshoot, adapt and recover. Nature 458:1102–1103. doi:10.1038/4581102a
Pierrehumbert RT (2014) Short-lived climate pollution. Annu Rev Earth Planet Sci 42:341–379. doi:10.1146/annurev-earth-060313-054843
Ramanathan V, Feng Y (2008) On avoiding dangerous anthropogenic interference with the climate system: formidable challenges ahead. Proc Natl Acad Sci USA 105:14245–14250. doi:10.1038/ngeo689
Ricke KL, Caldeira K (2014) Maximum warming occurs about one decade after a carbon dioxide emission. Environ Res Lett 9:1–8. doi:10.1088/1748-9326/9/12/124002
Seshadri AK (2015) Economic tradeoffs in mitigation, due to different atmospheric lifetimes of CO\(_2\) and black carbon. Ecol Econ 114:47–57. doi:10.1016/j.ecolecon.2015.03.004
Shindell DT, Lamarque J-F, Schulz M, Flanner M, Jiao C (2013) Radiative forcing in the ACCMIP historical and future climate simulations. Atmos Chem Phys 13:2939–2974. doi:10.5194/acp-13-2939-2013
Shoemaker JK, Schrag DP (2013) The danger of overvaluing methane’s influence on future climate change. Clim Change 120:903–914. doi:10.1007/s10584-013-0861-x
Skeie RB, Bernstsen T, Myhre G, Pedersen CA, Strom J (2011) Black carbon in the atmosphere and snow, from pre-industrial times until present. Atmos Chem Phys 11:6809–6836. doi:10.5194/acp-11-6809-2011
Skeie RB, Berntsen TK, Myhre G, Tanaka K, Kvalevag MM, Hoyle CR (2011) Anthropogenic radiative forcing time series from pre-industrial time until 2010. Atmos Chem Phys 11:11827–11857. doi:10.5194/acp-11-11827-2011
Smith SJ, Mizrahi A (2013) Near-term climate mitigation by short-lived forcers. Proc Nat Acad Sci USA 110(35):14202–14206. doi:10.1073/pnas.1308470110
Smith SJ, Wigley TML (2000a) Global warming potentials: 1. Climatic implications of emissions reductions. Clim Change 45:445–457. doi:10.1023/A:1005584914078
Smith SJ, Wigley TML (2000b) Global warming potentials: 2. Accuracy. Clim Change 45:449–469. doi:10.1023/A:1005537014987
Solomon S, Pierrehumbert RT, Matthews D, Daniel JS (2012) Atmospheric composition, irreversible climate change, and mitigation policy. In: Climate science for serving society: research, modelling and prediction priorities, Chap. Springer, New York
Stocker TF (2013) The closing door of climate targets. Science 339:280–282. doi:10.1126/science.1232468
Stouffer RJ (2004) Time scales of climate response. J Clim 17:209–217. doi:10.1175/1520-0442(2004)017<0209:TSOCR>2.0.CO;2
Williams KD, Ingram WJ, Gregory JM (2008) Time variation of effective climate sensitivity in GCMs. J Clim 23:2333–2344. doi:10.1175/2008JCLI2371.1
Winton M, Takahashi K, Held IM (2010) Importance of ocean heat uptake efficacy to transient climate change. J Clim 23:2333–2344. doi:10.1175/2009JCLI3139.1
Xie P, Vallis GK (2012) The passive and active nature of ocean heat uptake in idealized climate change experiments. Clim Dyn 38:667–684. doi:10.1007/s00382-011-1063-8
Zickfeld K, Eby M, Matthews HD, Weaver AJ (2009) Setting cumulative emissions targets to reduce the risk of dangerous climate change. Proc Natl Acad Sci USA 106:16129–16134. doi:10.1073/pnas.0805800106
Zickfeld K, Eby M, Weaver AJ, Alexander K, Crespin E, Edward NR (2013) Long-term climate change commitment and reversibility: an EMIC intercomparison. J Clim 26:5782–5809. doi:10.1175/JCLI-D-12-00584.1
Acknowledgments
This research has been supported by Divecha Centre for Climate Change, Indian Institute of Science. The author is grateful to several colleagues for helpful discussion. Two reviewers made suggestions that substantially improved the paper.
Author information
Authors and Affiliations
Corresponding author
Electronic supplementary material
Below is the link to the electronic supplementary material.
Appendices
Appendix 1: Solution of differential equations
The equation to be solved is
with initial condition \(\vec {u}\left( 0\right) =\vec {0}\). To solve we write the above equation as \(\dot{\vec {u}}-A\vec {u}=\vec {f}\), and multiply on the left by matrix exponential \({\text {e}}^{-At}\), yielding \({\text {e}}^{-At}\dot{\vec {u}}-{\text {e}}^{-At}A\vec {u}={\text {e}}^{-At}\vec {f}\), or
using identity \(\frac{{\text {d}}}{{\text {d}}t}{\text {e}}^{-At}=-Ae^{-At}=-{\text {e}}^{-At}A\) [see, for example, Hirsch and Smale (1974)]. The last equation is integrated for
using initial condition \(\vec {u}\left( 0\right) =\vec {0}\) and relation \(\left( {\text {e}}^{-At}\right) ^{-1}={\text {e}}^{At}\), which follows from the series expansion of \({\text {e}}^{At}\), i.e. \(I+At+\frac{\left( At\right) ^{2}}{2!}+\cdots\), where I is the identity matrix.
The matrix exponential in the solution above is simplified further by writing \(A=\varPhi \varLambda \varPhi ^{-1}\) with \(\varLambda\) the diagonal matrix of eigenvalues and \(\varPhi =\left[ \begin{array}{cc} \vec {\phi }_{1}&\vec {\phi }_{2}\end{array}\right]\) the \(2\times 2\) matrix containing the eigenvectors of A as columns. We know that \({\text {e}}^{At}=\varPhi {\text {e}}^{\varLambda t}\varPhi ^{-1}\) and \({\text {e}}^{-Az}=\varPhi {\text {e}}^{-\varLambda z}\varPhi ^{-1}\), which follow the series expansion of matrix \({\text {e}}^{At}\). Substituting this into Eq. (50) we obtain
Appendix 2: Approximation of fast contribution to global warming
Consider continuous function f(s) that is positive, and which can be approximated by a countable sequence of linear segments, so that \(\partial f/\partial s\) is constant in each segment, and with \(f\left( 0\right) =0\). Individual segments are indexed by \(i=0,1,....\) and \(\left[ s_{i},s_{i+1}\right]\) are their domains, with \(s_{0}=0\). Integration by parts shows that for coefficient \(c>0\)
equals
where \(\left( \frac{\partial f}{\partial z}\right) _{i}\) is the constant slope in segment i. Let \(\dot{f}_{{\text {sup}}}\equiv \sup _{i}\left( \frac{\partial f}{\partial z}\right) _{i}\) be the least-upper-bound of the segments’ slopes. Then
If, furthermore, \(s\gg 1/c\), then \({\text {e}}^{{\text {cs}}}-1\cong {\text {e}}^{{\text {cs}}}\). Under the condition that fluctuations in f(s) are slow compared to cf(s), so that \(\dot{f}_{{\text {sup}}}/c\ll f(s)\), we obtain
so that finally
If z denotes time, then c has units of inverse time. Integrating a positive function f(z) weighted by \({\text {e}}^{cz}\) is approximated by Eq. (56) if integration time s is large compared to 1 / c and, furthermore, growth of fluctuations over duration 1 / c is small compared to final value f(s). The analogy with physical systems is that linear systems close to equilibrium remain in equilibrium if forcing changes slowly. It is sufficient for the approximation above that the least-upper-bound of \(\partial f/\partial z\) be small enough, but in general the approximation would also be valid if fluctuations were small almost everywhere.
Returning to our problem, we treat forcing \(\tilde{F}(z)\) as a special case of f(z); and the negative of fast eigenvalue, \(-\lambda _{2}\), as a corresponding example of c above. Integration time is longer than 250 years, i.e. between simulation years 1765-2014, whereas \(-\lambda _{2}\, \gtrapprox \, 0.1\). Hence the first condition is met and \({\text {e}}^{-\lambda _{2}s}\gg 1\). The second condition is met if the change in radiative forcing within any 10-year period is much smaller than radiative forcing at the time of interest. With our focus on peak warming, where radiative forcing is still quite large, this condition is also met. Therefore, approximately \(\int _{0}^{t}{\text {e}}^{-\lambda _{2}z}\tilde{F}(z){{\text {d}}z}\cong \frac{{\text {e}}^{-\lambda _{2}t}\tilde{F}(t)}{-\lambda _{2}}\).
Appendix 3: Nonzero initial conditions
The solutions for the EBM in Sect. 3 are derived after assuming that initial conditions are zero, corresponding to preindustrial equilibrium. Here we relax this assumption, which is relevant to the case of pulse emissions. We start from Eq. (49)
but integrated from some time \(t_{0}\) where the state \(\vec {u}_{0}\ne \vec {0}\). Integrating
this is solved for
The effect of the forcing beginning at \(t=t_{0}\) is simply \({\text {e}}^{At}\int _{t_{0}}^{t}{\text {e}}^{-Az}\vec {f}(z){{\text {d}}z}\) or \(\varPhi {\text {e}}^{\varLambda t}\int _{t_{0}}^{t}{\text {e}}^{-\varLambda z}\varPhi ^{-1}\vec {f}(z){{\text {d}}z}\), which results in solutions derived in Sect. 3, with the only difference being that integration starts at \(t=t_{0}\). The EBM is linear, so its response to forcing is independent of the system’s state.
Rights and permissions
About this article
Cite this article
Seshadri, A.K. Fast–slow climate dynamics and peak global warming. Clim Dyn 48, 2235–2253 (2017). https://doi.org/10.1007/s00382-016-3202-8
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00382-016-3202-8