The current climate is the subject of intense debate because of the economic and societal implications of global warming. Greenhouse gases, mainly carbon dioxide but also other gases that are rarer but have a high absorption capacity of long-wave radiation re-emitted by the Earth, are accused of playing an important role in anthropogenic climate change. But the real anthropogenic impact on global temperature is difficult to assess accurately because of possible feedback loops and non-linear effects.
- Natural climatic cycles
- Spatial pattern of anthropogenic and natural temperature responses
- Regions primarily impacted by latent heat fluxes from the océans
- Regions primarily impacted by sensible heat fluxes from the océans
- Positive feedback
Yet the real anthropogenic contribution to climate change can be reliably known by taking into account the oceanic signature in the observed global surface temperature. This can be achieved by subtracting to the instrumental surface temperature the SST anomalies located on the internal antinodes of the Gyral Rossby Waves (GRW), that is, where the depth of the thermocline oscillates according to subharmonic modes. The oceanic contribution in the global temperature can be formally identified after 1870, the date from which the data is available, and before the anthropogenic impact becomes noticeable.
The impact on climate of the sea surface perturbation ΔT, which reflects the persistence of the vertical thermal gradient, either reinforces or, on the contrary, reduces evaporation. This results from atmospheric baroclinic instabilities that may lead to the formation of cyclonic systems. Because baroclinic instabilities of the atmosphere are most active when the resulting synoptic-scale eddies are stimulated and guided by the subtropical jet streams, SST anomalies are located at high latitudes of the gyres around 40°. They have to be representative of long-period GRWs, which correspond to high subharmonic modes, to accurately reflect the persistence of their continental replicas.
Consequently, areas representative of the oceanic signature in the global temperature are to be selected at high latitudes of the five subtropical gyres. The perturbation ΔT represented by the SST anomalies, averaged over each area, is deduced from the land surface temperature. The resulting oceanic signature of the global temperature is obtained by doing the weighted average of the SST anomalies in the five subtropical gyres. The weights are indicative of the incidence on the global temperature of the corresponding gyres, that is, they are approximately proportional to the areas of the continents impacted by each of the gyres.
Some continental regions are directly impacted by heat exchanges between the oceans and the continents. To these continental regions can be associated particular SST anomalies by jointly analyzing, both in space and time, the SST and the rainfall oscillations in the 5–10 year band. This method allows representing the inland areas subject to such rainfall oscillation, from which the signatures of the SST and rainfall height anomalies can be unambiguously associated. This is made possible because of the selectivity of both SST and rainfall height anomalies within this band. The active SST anomalies are located on the internal antinodes of the GRWs for the relevant subharmonic modes. The main areas subject to rainfall oscillation at mid-latitudes where condensation / precipitation of water vapor releases the latent heat, are Southwest North America, Texas, Southeastern and Northeastern North America, Southern Greenland, Central and Western Europe and Western Asia, the region of the Río de la Plata, Southwestern and Southeastern Australia, and Southeast Asia.
However, SST anomalies in the 5–10 year band are representative of short-term exchanges between the oceans and the continents resulting from the resonance of GRWs for low subharmonic modes and their inland thermal imprints are evanescent. For these reasons, representative areas of SST anomalies on the long-period internal antinodes, which correspond to higher subharmonic modes, must be judiciously selected to accurately represent the persistence of continental thermal footprints. Accuracy of the SST anomalies averaged over such areas requires the latter are as small as possible not to integrate short-term exchanges the signature of which is masked by long-term exchanges. Short-term and long-term exchanges are governed by short-wavelength and long-wavelength GRWs, respectively. Internal antinodes of short-wavelength GRWs extend from where the western boundary current leaves the coast to the bifurcation of the re-circulating wind-driven current of the gyre and the drift current leaving the gyre. Internal antinodes of long-wavelength GRWs extend all around the gyres so that areas representative of persistent exchanges are necessarily located to the east of the short-period internal antinodes.
Preselected areas are considered as representative of thermal exchanges in the perturbed state of the global climate system when the oceanic perturbation ΔT, that is, the weighted average of SST anomalies over the five subtropical gyres, is a replica of the instrumental global temperature. This can be done before the global temperature is subject to anthropogenic warming. Then, the net anthropogenic contribution in the global temperature can be estimated by subtracting from the latter the weighted sum of the SST anomalies from the global temperature. Actually, the contributions of the SST anomalies are estimated by using the least squares method, that is, by minimizing the sum of the squares of the differences between the instrumental global temperature and the weighted sum of the SST anomalies within a relevant interval of time, the sum of the weighting factors being one.
Oceanic signatures exhibit particular behaviors according to the gyres. In the Figure the instrumental surface temperature is compared to the weighted sum of SST anomalies wNANA+wNPNP+wSASA+wSPSP+wSISI where the weighting factors are wNA=0.50, wNP=0.17, wSA=0.15, wSP=0.13, wSI=0.05. Systematic differences are observed. Beyond 1970, the discrepancies highlight the contribution of the anthropogenic warming. Before 1900 they reflect systematic errors on measurements.
The contribution of the component in the band 48-96 years, whose amplitude of variation is 0.3°C, is significant as well as that in the band 192-576 years, which varies between ±0.1°C. The latter can be considered as a rebound following the little ice age although this subharmonic mode is weakly exogenously forced and behaves as a harmonic of lower frequency GRWs. The 256-year average period GRWs are indeed coupled to those of 128-year average period, which are forced by the Gleissberg cycle of the Sun.
The TSI reconstructed is decomposed into the frequency bands representative of the subharmonic modes, which allows the accurate estimation of the forcing efficiency in the 96-192 year band. Calculated from the maximum oscillation occurring in 1976 for both the TSI and the global temperature, it is 0.21 °C(W/m2)-1. This estimate is very low compared to what happens during the Holocene and corresponds to a low amplitude of the Gleissberg cycle that occurs after 700 years BP and more particularly 300 years BP. To compare, this forcing efficiency is close to the value deduced from the greenhouse effect resulting from the increase in atmospheric water vapor following an increase in the global temperature, that is, 0.22 °C(W/m2)-1.
The weights associated with the SST anomalies that represent at best the gridded surface temperatures Ts are estimated by using the same least squares method as that explained previously. Here, the time interval from which the fitting is performed is 1940-1970, for which the weights are the most precise and the most representative of natural forcing when the surface temperatures are considered individually in the 5°×5° grid. This choice allows us to minimize the noise in the spatial pattern of the natural temperature. However the estimation of the part of the anthropogenic response within Ts by subtracting from the latter the weighted sum of the SST anomalies very little depends on the time interval, 1900-1970 or 1940-1970.
The temperature response to the natural radiative forcing exhibits a low spatial variability in both hemispheres. In the northern hemisphere it is because the temperature response of the Atlantic and the Pacific oceans in 2015 are close (the temperature increase since 1970 in the Pacific is slightly lower than in the Atlantic). The influence of the Pacific can be seen in the Central Asia whereas North America and Europe are rather influenced by the Atlantic. The southern hemisphere reflects the influence of the warmer Indian Ocean than the other two oceans. Everywhere the natural temperature response is positive because all oceanic signatures increase since 1970s. The increase is most noticeable in the North America and north of 60°N, where it reaches 0.6 ° C.
In addition to the natural response, the part of the anthropogenic response within the instrumental surface temperature Ts shows considerable spatial variability. Lower than 0.8°C and even 0.5°C in Australia, southern South America, eastern North America, northern and Western Europe, and Southeast Asia it overreaches 2°C in Eastern Europe, Russia, Kazakhstan, Mongolia, east of North America, east of Brazil, eastern Africa, Angola, Namibia, even more than 2.5°C north of 70°N. This great disparity questions the nature of the positive feedback loop responsible for such amplification in some regions, regardless the latitude.
The distribution of extra-tropical regions with low anthropogenic impact coincides with those subject to the rainfall oscillation in the 5-10 year band. In the North American continent, these are mainly the regions of eastern and south-western United States. In South America this concerns the countries of the north, and both eastern and southern Argentina. In Europe, these are the South Greenland, and western and northern countries. In Africa, the oscillation concerns the North of the Maghreb countries and South Africa. In Oceania the oscillation is observable almost everywhere.
At mid-latitudes, the 5–10 year band rainfall oscillation characterizes regions impacted by latent heat transfers from the oceans. Within the intertropical convergence zone, monsoonal regions are weakly impacted by anthropogenic warming, that is, the Central America, Western Africa, India and the South-East Asia.
The areas that are heavily impacted by anthropogenic warming are characterized by a high amplitude of the rainfall oscillation in the 0.5-1.5 year band, exhibiting a strong seasonality. The annual rainfall pattern displays a peak time in late boreal and austral summer, that is, when the difference between the temperature of the air aloft and the surface temperature is the greatest, leading to the greatest potential for instability. In this way the precipitations occur endogenously within the continents, and mainly sensible heat transfer occurs from the oceans.
Amplifying effects on the anthropogenic temperature response do not depend on the latitude, but on the way in which the thermal exchanges occur between the oceans and the continents. The south Greenland and the eastern Africa are respectively little and heavily impacted, which is contrary to the general trend. On the other hand, contrary to commonly accepted ideas, amplifying effects do not result from the greenhouse effect resulting from the supposed increased atmospheric water vapor associated with the increased temperature. Regions least affected by anthropogenic warming are such that the increase in water vapor with temperature occurs in the free troposphere.
In regions impacted by latent heat fluxes extra-tropical free tropospheric water vapor (above the boundary layer) is mostly associated with the tropospheric circulation. An increase in the temperature of the atmosphere increases its water-holding capacity, in proportion with the Clausius-Clapeyron relation. In contrast, arising from endogenous primary moisture sources, free tropospheric water vapor in regions impacted by sensible heat fluxes is mainly controlled by the difference between the temperature of the air aloft and the surface temperature.
Those findings reinforce the idea that the climate response is closely linked to the top of atmosphere flux as suggested by the spatial pattern of climate feedback. The only way indeed to explain the spatial distribution of the anthropogenic temperature response is to assign a driving role in the amplification effect of the high troposphere cloud cover, which involves the lapse rate, surface albedo and cloud feedbacks.
Jet-streams are fast winds aloft blowing from west to east. Along a curved and sinuous path, they play a major role in atmospheric circulation as they participate in the formation of depressions and anticyclones at middle latitudes, which then move under these powerful atmospheric currents.
Positive feedback loops amplify changes in a dynamical system; this tends to move the system away from its equilibrium state and make it more unstable. Negative feedbacks tend to dampen changes; this tends to hold the system to some equilibrium state making it more stable.