57(Bony et al. 2006 ) and transport of heat from the atmosphere to the ocean that drives
a long term rise in the temperature of the world’s oceans (Levitus et al. 2012 ).
Our Empirical Model of Global Climate (EM-GC) (Canty et al. 2013 ) uses an
approach termed multiple linear regression (MLR) to simulated observed monthly
variations in the global mean surface temperature anomaly (termed ΔTi, where i is
an index representing month) using an equation that represents the various natural
and anthropogenic factors that influence ΔTi. The EM-GC formulation represents:
- RF of climate due to anthropogenic GHGs, tropospheric aerosols, and land use
change - Exchange of heat between the atmosphere and ocean, in the tropical Pacific,
regulated by ENSO - Variations in TSI reaching Earth due to the 11 year solar cycle
- Reflection of sunlight by volcanic aerosols in the stratosphere, following major
eruptions - Exchange of heat with the ocean due to variations in the strength of AMOC
- Export of heat from the atmosphere to the ocean that causes a steady long-term
rise of water temperature throughout the world’s oceans
The effects on ΔT of the Pacific Decadal Oscillation (PDO) (Zhang et al. 1997 ) and
the Indian Ocean Dipole (IOD) (Saji et al. 1999 ) are also considered.
The hallmark of the MLR approach is that coefficients that represent the impact
of GHGs, tropospheric aerosols, ENSO, major volcanoes, etc. on ΔTi are found, such
that the output of the EM-GC equations provide a good fit to the observed climate
record. The most important model parameters are the total climate feedback param-
eter (designated λ) and a coefficient that represents the efficiency of the long- term
export of heat from the atmosphere to the world’s oceans (designated κ). Our
approach is similar to many prior published studies, including Lean and Rind ( 2009 ),
Chylek et al. ( 2014 ), Masters ( 2014 ), and Stern and Kaufmann ( 2014 ) except ocean
heat export (OHE, the transfer of heat from the atmosphere to the ocean) is explicitly
considered and results are presented for a wide range of model possibilities that pro-
vide reasonably good fit to the climate record, rather than relying on a single best fit.
Most of the prior studies neglect OHE and typically rely on a best fit approach.
A description of the EM-GC approach is provided in the remainder of this section.
While we have limited the use of equations throughout the book, they are necessary
when providing a description of the model. We’ve concentrated the use of equations
in the section that follows; comparisons of output from the EM-GC with results from
the CMIP5 GCMs are presented in other sections with use of little or no equations.
2.2.1 Formulation
The Empirical Model of Global Climate (Canty et al. 2013 ) provides a mathemati-
cal description of observed temperature. As noted above, temperature is influenced
by a variety of human and natural factors. Our approach is to compute, from the
historical climate record, numerical values of the strength of climate feedback and
2.2 Empirical Model of Global Climate