Global land-surface warming much faster than ocean surface, and Northern Hemisphere faster than Southern: incriminates soot from burning oil and coal, exonerates CO2

Fossil-fuel combustion now outweighs solar variations in driving climate change (Higgs 2022, GSA, www.researchgate.net/publication/362103181). Remarkably, land (near-surface-air) warming is three times faster than ocean-surface warming, and Northern Hemisphere (NH; land-ocean average) three times Southern. This aberrant behavior began abruptly in 1985 (contrast pre-1985 lockstep warming-cooling; data.giss.nasa.gov/gistemp/graphs_v4). Moreover, over land and over the NH, warming is significantly slower at altitude (UAH satellite-measured lower-troposphere average temperature).

These strong lateral- and vertical warming gradients incriminate airborne soot (warms atmosphere by absorbing solar radiation). Soot’s poor dispersal causes strong concentration gradients, both (A) laterally, toward its main sources, which are predominantly on-land and NH (diesel engines, cooking woodfires, coal-fired powerplants/industries), e.g. over intensely industrialized nations (USA, Europe, China, etc.), average atmospheric soot concentration is ~1000% (i.e. 10 times) greater than over adjacent oceans (NASA 2011 global black carbon video https://svs.gsfc.nasa.gov/3844), starkly contrasting with CO2’s 1% difference (NASA global CO2 video); and (B) vertically, e.g. year-round average soot concentration above rural Siberia is ~500% higher at 0.5km than at 3km (doi: 10.3390/atmos12030351), far exceeding CO2’s 4% difference above Tokyo (10.3390/s18114064).

Two further observations implicate diesel- and coal-sourced soot specifically. Firstly, 25 years (y) before the 1985 decouplings (above), world annual oil consumption tripled in 1960, then remained high almost continuously (OurWorldinData, GlobalPrimaryEnergyConsumptionBySource graph). Secondly, coal’s distinctively stepwise growth (same graph) is mimicked, with a similar time-lag (10-20y), by stepwise land-air warming (data.giss.nasa.gov/gistemp/graphs_v4): COAL GROWTH fast 1974-1989 (tripled in 1974, due to 1973 oil crisis), nil 1989-1999, fast 1999-2014 (mainly China; OurWorldinData, CoalConsumptionByRegion graph); LAND WARMING fast 1994-2005, nil 2005-2011 (famed ‘hiatus’), fast since 2011.

CO2 cannot explain the observed strong lateral and vertical warming gradients, because its efficient dispersal produces near-homogenous atmospheric concentration. Even heavily industrial regions barely (<0.5%) exceed the global average (10.1038/s41598-019-53513-7). Furthermore, no leap in CO2 concentration occurred ~1985 or any other time; instead, CO2 grew by gradual acceleration, not stepwise (keelingcurve.ucsd.edu). Evidently, CO2 has negligible effect on climate, implying that its greenhouse effect is nullified by unknown and/or underestimated feedbacks (e.g. 10.1007/978-94-007-6606-8_17). If so, hyper-expensive CO2 capture is misconceived, besides counter-productive (today’s 420ppm is well below ~1,000ppm optimum for crop- and forest growth).

In the literature, the global-warming contribution of soot (‘black carbon’) is very uncertain. According to an influential review (10.1002/jgrd.50171; italic emphasis added here): “The best estimate of industrial-era climate forcing of black carbon ... is +1.1 W m^-2 with 90% uncertainty bounds of +0.17 to +2.1 W m^-2 (sic) ... We estimate that black carbon ... is the second most important human emission in terms of its climate forcing”. Black carbon’s warming effect was estimated to be 70% as strong as CO2. Recent IPCC estimates are 35% and 12% (2013, Physical Science Basis, Summary for Policymakers, fig.SPM.5; 2021, ditto, fig.SPM.2c). On the contrary, the data presented above suggest black carbon is overwhelmingly the dominant anthropogenic-warming agent. Helping to explain previous underestimates, two additional soot-induced warming mechanisms, via its effects on clouds, were recently recognised (10.1038/s41561-020-0631-0). Moreover, developing-world powerplants possibly emit far more soot (10.1029/1999JD900187) than the review assumed.