Dr. Lean and her colleagues used data on climate – surface temperature, ENSO, volcanoes, solar variability and anthropogenic climate factors to construct an empirically based multi factor, linear climate model.
Climate is far from linear in any sense. And numerical climate models use the same equations of fluid motion that Edward Lorenz used in his 1960’s convection model - to discover chaos theory. So climate models are intrinsically chaotic as well. But an empirical, multi factor analysis might at least give some insight into the modes of major climate variation. The top of atmosphere (TOA) radiant flux, the radiant imbalance at any time, ultimately both drives climate and changes at the speed of light – rather than lagging as global tropospheric temperature does from coupled ocean and atmosphere processes.
Changes in the spectral adsorption characteristics of the atmosphere appear in radiant flux at the top of atmosphere (TOA) where energy is – in the deepest sense of theoretical physics – eventually in radiant equilibrium. Over an interval - the incoming energy (Ein) must equal the outgoing (Eout) plus the change in ocean heat storage (OHS). I give you the differential equation of global energy storage.
Ein = Eout + (GES t+1 – GESt)/∆t = Eout + d(GES)/dt
Energy in and out is the measured energy flux (usually reported as Watts per metre squared – W/m2) at TOA over the interval and is in Joules (J) - or Joules per metre squared (J/m2). One Watt for one second is one Joule. Global heat storage is almost entirely in the oceans and is measured by the ARGO project - thousands of probes measuring temperature and salinity in the oceans since 2003. Heat energy in oceans in Figure 4 is given on Joules per metre squared (J/m2).
Figure 1 shows International Satellite Cloud Climate Project (ISCCP – FD) monthly long wave (LW) radiant flux. Eyeballing the data - there is a net increase in LW radiant flux up at TOA. 
Figure 1: ISCCP FD – global mean long wave radiant flux up at TOA
The trend with greenhouse gases and water vapour is to lower upward LW flux. Energy is re-emitted much more strongly from a warmer atmosphere than a cooler – a 4th order exponential function – however, LW up at TOA might be expected to decline as the LW flux is changed by anthropogenic influences in the atmosphere. But the LW trend is to increased upward flux. There must be at least one other factor in the Earth LW radiant flux. Clouds have a role in LW flux. Less cloud equals less adsorption and more LW up. Does this show decadal – between the 1980’s and 1990’s – change in cloudiness which more than offset the anthropogenic LW changes?
A decrease in cloudiness is confirmed by the short wave (SW) record. Decreasing reflectance - from primarily less cloud cover - to 2000. The cooling influence of volcanic aerosols from El Chichón in the early 80’s and Mt Pinatubo in the early 90’s can be seen. SW fluxes upward and out of the Earth energy system declined as cloud cover declined from the 1980’s through to 2000. 
Figure 2:ISCCP FD – global mean short wave radiant flux at TOA
Cloud cover changes are energetically important in the tropics and subtropics. ‘Tropical and subtropical low-level marine clouds consist of optically thick stratocumulus clouds, which usually form over the regions associated with relatively cold sea surface temperatures (SSTs) and the atmospheric subsidence, and optically thin shallow cumuli in the tradewind regime. These low-level clouds play a pivotal role in the global climate system not only by affecting radiative budgets but also by promoting heat and moisture exchange between the sea-surface, the boundary layer, and the overlying troposphere.’
Ping Zhu and colleagues - Climate sensitivity of tropical and subtropical marine low cloud amount to ENSO and global warming due to doubled CO2 - JGR, VOL. 112, 2007
The planet gained energy since 2000 in the NASA’s ‘Clouds and the Earth Radiant Energy System’ (CERES) data in Figure 3. Net radiant flux, by convention, shows the planet gaining and losing energy. The more positive the change the faster the rate of planetary warming.
The net radiative flux increase over the period shows a planetary warming trend associated with a modest decrease in reflected SW radiant flux from 2000 to 2006 and little change since. AGW theory would suggest a - 0.37 W/m2/decade anthropogenic trend in LW. The change in LW is less even than that in reflected SW decrease.
TOA radiant flux, LW out, SW out and Net in – W/m2
Figure 3: CERES LW, SW and NET TOA radiant flux
Oceans are warmed up to 100 m deep by incoming SW. Warm water is mixed in turbulent ocean currents to great depth. Land surfaces gain and lose heat in the diurnal and seasonal cycles. Most of the long term net SW energy gain shows up in ocean heat storage. Marginally less cloud and more SW heating the ocean should drive modest warming in the deep oceans – as Karen Von Schuckmann and colleagues show from ARGO data. The net warming from SW changes does seem to be showing up as expected in deep ocean heat storage.
Figure 4: Deep ocean heat storage (Von Schuckmann et al, 2009)
The ‘radiant imbalance’ in the shortwave is not showing directly in the tropospheric temperature record – which at any rate is influenced by coupled ocean and atmosphere processes. Global tropospheric temperature peaks, in this period, in 2002.
Figure 5: UAH Global tropospheric temperature anomaly
ENSOhas a large influence on interannual to decadal tropospheric temperature variations. Conditions changed, in the ‘Great Pacific Climate Shift’ of 1976/1977, to a warm Pacific SST mode and to marginally cooler again in 1998/2002. Decadal variations show up in changes in wind, SST and cloud across the Pacific in the Wolter MEI below. There was a tendency to La Niña (blue) conditions prior to 1977 and after 1998 – with El Niño (red) conditions dominant in 1977 to 1998. There is a relatively lower SST in La Niña - and relatively lower tropospheric temperature as in 2000 and 2008.
Figure 6: Multivariate ENSO index
CERES data since 2000 clearly show SW changes – this drove the ocean heat content increase measured in ARGO. Most of recent warming is driven by changes in reflected SW. SW changes imply changes in global cloud cover over the period. Cloud formation depends on sea surface temperature (SST) – an ENSO feedback – which varies over years and decades. With more upwelling and cooler water in the Pacific for a decade or two more – in the current cool Pacific decadal variation (PDV) – colder SST give relatively more cloud and suggest the possibility of decadal shading of atmosphere and oceans.
Over the next decade or two the chances are that global surface temperature changes will be modest in the current cool PVD. Beyond that the complex dynamic of Earth climate - clouds, ice, rain and snow, biology, landform, dust and aerosols, UV warming of ozone in the stratosphere, deep water formation and deep ocean upwelling – is certain to include surprises. Abrupt climate change occurs globally in interannual to multidecadal and millennial increments. Climate has changed ‘dangerously’ in the past in a matter of months to years – and many smaller and less persistent abrupt changes can be seen in climate records. A 2002 US National Academy of Science report, ‘Abrupt climate change: inevitable surprises’ addresses abrupt climate change in both paleoclimate proxies and modern climate records.
Oceanographer Wally Broecker has likened anthropogenic greenhouse gas emissions to poking a stick at a dangerous beast. Something a child would do. But scientific understanding evolves with data and theory testing. With large gaps in knowledge - even as to the sign of theoretical global warming cloud feedback - it should really be enough for science to identify something we shouldn’t be doing to get on with figuring the best way to curb the behaviour or effect.
