How NOAA and Bad Modeling Invented an “Ocean Acidification” Icon: Part 2 – Bad Models

Guest essay by Jim Steele

Director emeritus Sierra Nevada Field Campus, San Francisco State University and author of Landscapes & Cycles: An Environmentalist’s Journey to Climate Skepticism

Are the Oceans’ Upper Layers Really Acidifying?

As detailed in Part 1, NOAA’s Bednarsek, incorrectly attributed the dissolution of sea butterfly shells to anthropogenic CO2 although the evidence clearly showed the natural upwelling of deeper low pH waters were to blame. Based on models employed by NOAA’s Feely and Sabine, Bednarsek claimed the upper ocean layers are becoming more acidic and less hospitable to sea butterflies relative to pre-industrial times. However detecting the location and the depth at which anthropogenic CO2 now resides is a very, very difficult task. Because the ocean contains a large reservoir of inorganic carbon, 50 times greater than the atmospheric reservoir, the anthropogenic contribution is relatively small. Furthermore anthropogenic carbon comprises less that 2% of the combined CO2 entering and leaving the ocean surface each year. Thus there is a very small signal to noise ratio prohibiting accurate detection of anthropogenic CO2. Despite admittedly large uncertainties, modelers boldly attempt to infer which layers of the ocean are acidifying.

(To clarifying terminology, an organic carbon molecule is a molecule that is joined to one or more other carbons, such as carbohydrates and hydrocarbons. CO₂ with a lone carbon is considered inorganic, and when dissolved can take 3 forms (or “species”) collectively referred to as Dissolved Inorganic Carbon (henceforth DIC): 1) Carbonic acid (H₂CO₃), 2) Bicarbonate ion (HCO₃^–) after losing one H+ 3) Carbonate ion (CO₃^-2) after losing a second H+ )

However model results are based on three very dubious assumptions:

1) Models assume surface layers absorb anthropogenic CO2 by reaching equilibrium with atmospheric concentrations. With minor adjustments, models simply calculate how much dissolved inorganic carbon (DIC) will be added to the ocean based on increased atmospheric CO2 since pre-industrial times.

2) Models assume CO2 will diffuse into the upper ocean layers and be transported throughout the ocean in a similar fashion to tracers, like CFCs. Because CFCs accumulate disproportionately near the surface, models assume DIC does as well.

3) Models assume biosphere is in a steady state. Thus they do not take into account increased primary production and the rapid export of carbon to depth.

Although there is no doubt anthropogenic CO2 is taken up by the oceans, assertions that ocean surface layers are acidifying are the results of faulty model assumptions.

What Equilibrium?

CO2 equilibrium is rarely achieved between ocean and atmosphere. Ocean surface pH and thus calcium carbonate saturation levels are determined by the efficiency of the biological pump. In other words, when, where, and how much CO2 enters the ocean surface, requires surface CO2 concentrations to be lower than atmospheric concentrations. That difference depends on how much CO2 is fixed into organic carbon by photosynthesis and subsequently exported it to depth, or how much CO2 is upwelling. Photosynthesis indiscriminately draws down all CO2 molecules that have invaded surface waters either via upwelling from depth or by diffusion from the atmosphere. Despite opposing effects of mixing and diffusion, the biological pump maintains a strong vertical gradient of high surface water pH and low DIC, with decreasing pH and increasing DIC at greater depths. In regions where strong upwelling of DIC from the deeper ocean overwhelms the ability of photosynthesizing organisms to sequester carbon, surface pH drops and CO2 is outgassed to the atmosphere. Several models estimate that without the biological pump, atmospheric CO2 would increase by 200 to 300 ppm above current levels.

The efficiency of the biological pump determines to what depths anthropogenic carbon will be transported. However NOAA’s Sabine does not model the effects of the biological pump, oddly stating “although ocean biology plays an integral role in the natural distribution of carbon in the ocean, there is no conclusive evidence that the ocean uptake and storage of anthropogenic carbon, thus far, involve anything other than a chemical and physical response to rising atmospheric CO2.”

Does Sabine truly believe the undeniable biological pump discriminates between anthropogenic and natural carbon? Or does he believe that there have been no changes in primary production and carbon export? As primary production increases, so does the carbon export to depth. Annual primary production in the Arctic has increased by 30% since 1998. We can infer primary production increased in the Sargasso Sea based on a 61% increase in mesoplankton between 1994 and 2006. North Atlantic coccolithophores have increased by 37% between 1990 and 2012. And primary production and carbon export in the Peru Current has dramatically increased since the end of the Little Ice Age. The increasing trend in primary production and accompanying carbon export is potent evidence supporting an alternative hypothesis that the biological pump has sequestered increased invasions of anthropogenic CO2.

An examination of seasonal changes in surface CO2 concentration, illustrates how the biological pump determines when and how much CO2 enters the ocean, and how much DIC accumulates near the surface. As exemplified by the graph below from 2008 buoy data off the coast of Newport, Oregon (Evans 2011), each spring photosynthesis lowers ocean surface CO2 to 200 ppm, far below current atmospheric concentrations and much lower than what would be expected from equilibrium with a pre-industrial atmosphere. Spring surface waters are supersaturated, and any downwelling or mixing of these supersaturated waters cannot acidify upwelled water or subsurface layers. Furthermore the springtime draw down conclusively removes any anthropogenic CO2 residing in sunlit waters. Furthermore experiments show CO2 is often a limiting nutrient and added atmospheric CO2 stimulates photosynthesis. Microcosm experiments found that when atmospheric CO2 was increased, the plankton community consumed 39% more DIC.

Upwelling season begins in summer extending through fall. As illustrated above, upwelling events rapidly raise surface concentrations of CO2, reaching 1000 ppm. Physics dictates there can be no net diffusion from the atmosphere into the ocean when the oceanic concentration is higher than atmospheric, and thus there are virtually no anthropogenic additions during upwelling season. Here any lowering of surface pH or calcium carbonate saturation must be due to upwelling.

Finally during the winter, (not illustrated) surface waters exhibited a steady CO2 concentration of 340 ppm. Although photosynthesis is reduced, and winter mixing brings more subsurface carbon and nutrients to the surface, the surface remains below equilibrium with the atmosphere. Although surface concentrations are low enough to permit the invasion of atmospheric CO2, the biological pump continues to export that carbon to depth so that surface layers remain supersaturated all winter.

Diffusion of CO2 into the ocean is a slow process. It is believed that it requires about 1 year for the oceans to equilibrate with an atmospheric disturbance. But as spring arrives, increasing sunlight again enhances photosynthesis, so whatever anthropogenic CO2 that may have invaded the surface over the course of the year, is once again fully sequestered and pumped to depth, lowering surface CO2 concentrations to 200 ppm. Bednarsek’s claim that anthropogenic CO2 is acidifying the upwelled water along the Oregon California Coast is once again not supported.

Tracers Do Not Correctly Simulate Transport of Anthropogenic Carbon

Tracers like chlorofluorocarbons (CFCs) are synthetic gases that are biologically inert. They were introduced to the world during the 1920s primarily as a refrigerant. Climate scientists have assumed the physical transport and accumulation of CFCs and increasing anthropogenic carbon will be similar. Below in Figure 1, the red area just south of Greenland designates an area that has accumulated the most CFCs. This local concentration happens when high salinity Atlantic waters cool and carry surface water and its dissolved gasses downward to the abyss forming North Atlantic Deep Water. It is estimated that this downwelling has exported 18% of all CFCs below 1000 meters; implying dissolved anthropogenic carbon has been similarly exported and sequestered. However elsewhere CFCs accumulate disproportionally in upper surface layers, so models assume dissolved anthropogenic CO2 is likewise accumulating nearer the surface.

Both CFCs and CO2 are gases, and their solubility is similarly modulated by temperature. Warm waters of the tropics absorb the least amount of CFCs and CO2, as illustrated by the dark blue regions in Figure 1 from Willey 2004. Thus equatorial waters feeding the California Undercurrent that upwell along the west coast have likewise absorbed the least amounts of anthropogenic carbon, if any. (The extremely low level of CO2 diffusion into the tropical ocean plus the super saturation of tropical waters, casts great misgivings on any claim that coral reefs have been significantly affected by anthropogenic acidification.)

However, unlike inert CFCs, any CO2 entering sunlit waters is quickly converted to heavy organic matter by photosynthesis. Although dissolved CFCs and dissolved carbon are passively transported in the same manner, particulate organic carbon (alive or dead) behaves very differently. Particulate carbon rapidly sinks, removing carbon from the surface to depth in ways CFC tracers fail to simulate. Examination of the literature suggests “various methods and measurements have produced estimates of sinking velocities for organic particles that span a huge range of 5 to 2700 meters per day, but that commonly lie between tens to a few hundred of meters per day”. Low estimates are biased by suspended particles that are averaged with sinking particles. Faster sinking rates are observed for pteropod shells, foraminifera, diatoms, coccolithophorids, zooplankton carapaces and feces aggregations, etc that are all capable of sinking 500 to 1000 meters per day. These sinking rates are much too rapid to allow respired CO2 from their decomposition to acidify either the source waters of upwelling such as along the Oregon and California coast, or the surface waters

Earlier experiments had suggested single cells sank very slowly at rates of only 1 meter per day and thus grossly underestimated carbon export. However single-cell organisms will aggregate into clusters that increase their sinking rates. Recent studies revealed the “ubiquitous presence of healthy photosynthetic cells, dominated by diatoms, down to 4,000 m.” Based on the length of time healthy photosynthesizing cells remain viable in the dark, sinking rates are calculated to vary from 124 to 732 meters per day, consistent with a highly efficient biological pump. Although NOAA’s scientists have expressed concern that global warming will reduce the efficiency of the biological pump by shifting the constituents of phytoplankton communities to small, slow-sinking bacteria, new research determined that bacteria also aggregate into clusters with rapid sinking rates ranging from 440 to 660 meter per day.

Sequestration of carbon depends on sinking velocities and how rapidly organic matter is decomposed. Sequestration varies in part due to variations in the phytoplankton communities. Depths of 1000 meter are considered to sequester carbon relatively permanently, as waters at those depths do not recycle to the surface for 1000 years. Weber 2016 suggests 25% of the particulate organic matter sinks to 1000 meter depths in high latitudes while only 5% reaches those depths at low latitudes. But long-term sequestration does not require sinking to 1000 meter depths. Long-term sequestration requires sinking below the pycnocline, a region where the density changes rapidly. Dense waters are not easily raised above the pycnocline, so vertical transport of nutrients and carbon is inhibited creating long-term sequestration. Because the pycnocline varies across the globe, so do sequestration depths.

Below on the left is a map (a) from Weber 2016, estimating to what depths particles must sink in order to be sequestered for 100 years. Throughout most of the Pacific particles need only sink to depths ranging from 200 to 500 meters. In contrast the golden regions around the Gulf Stream, New Zealand and southern Africa must sink to 900 meters.

The map on the right (b), estimates what proportion of organic matter leaving the sunlit waters will be sequestered. The gold in the Indian Ocean estimates 80% will reach the 100-year sequestration depth, while 60% will reach sequestration depths along the Oregon California Coast. Again casting doubt on Bednarsek’s claims of more recently acidified upwelled waters acidifying sea butterfly shells. Elsewhere in map “b”, 20% or less of the exported carbon reaches sequestration depths.

The combination of sinking velocities and sequestration depths suggests significant proportions of primary production will be sequestered in a matter of days to weeks. This is consistent with the maintenance of the vertical DIC and pH gradients detected throughout our oceans. However it conflicts with claims by NOAA’s scientists.

Biased by CFC observations Sabine wrote, “Because anthropogenic CO2 invades the ocean by gas exchange across the air-sea interface, the highest concentrations of anthropogenic CO2 are found in near-surface waters. Away from deep water formation regions, the time scales for mixing of near-surface waters downward into the deep ocean can be centuries, and as of the mid-1990s, the anthropogenic CO2 concentration in most of the deep ocean remained below the detection limit for the delta C* technique.”

That NOAA scientists fail to incorporate the fact that particulate carbon can be sequestered to harmless depths in a matter of days to weeks instead of “centuries” appears to be the cause of their catastrophic beliefs about ocean acidification. Furthermore because CFCs have accumulated near the surface with only miniscule amounts in the deeper ocean, the tracer provides absolutely no indication of how upwelling brings ancient DIC to the surface. So by relying on a CFC tracer, their models will mistakenly assume that increased concentrations of DIC near the surface must be due to accumulating anthropogenic carbon, and not upwelled ancient carbon.

The Ocean’s Biosphere Steady State?

Given a steady export percentage of primary productivity, increasing amounts carbon will be exported in proportion to increasing productivity. Thus it is reasonable to hypothesize that if marine productivity has increased since the end of the Little Ice Age (LIA) aka pre-industrial times, that increased production will have sequestered the increasing amounts of anthropogenic carbon. Although there are only a few anoxic (depleted oxygen) ocean basins, where organic sediments can be well preserved, those basins all reveal that since the Little Ice Age, marine productivity and carbon export has indeed increased as the oceans warmed.

Research from Chavez 2011 is illustrated below and demonstrates that during the LIA, marine primary productivity (d) was low, but has increased by 2 to 3 fold over the recent 150 years. Sediments reveal that fast sinking diatoms increased 10 fold at the end of the LIA, but likely due to silica limitations, since 1920 diatom flux to ocean sediments has been reduced to about a 2-fold increase over the LIA. Nonetheless numerous studies find estimates of sedimentary diatom abundance is representative of carbon export production in coastal upwelling regions.

The increased primary production coincides with over a hundred-fold increase in fish scales and bones (f). And consistent with the need for increased nutrients to support increased primary production, proxy evidence suggests a 2-fold increase in nutrients in the water column (c). Such evidence is why researchers have suggested their observed decadal increases in upwelled DIC and nutrients might be part of a much longer trend. Finally in contrast to the global warming explanation for depleted ocean oxygen, the decomposition of increased organic carbon provides a more likely explanation for observations of decreased oxygen concentrations in the water column (a) and sediments (b). Because primary production had doubled by 1900, long before global warming or before anthropogenic CO2 had reach significant concentrations, it is unlikely anthropogenic CO2 contributed to increased upwelling, increased primary production or any other trends in this region

However increased primary production alone does not guarantee that sinking particulate carbon is removing enough carbon to counteract anthropogenic additions. However there are dynamics that suggest this must be the case. First consider that an examination of the elements constituting a phytoplankton community, there is a common ratio of 106 carbon atoms detected for every 16 nitrogen atoms (aka a Redfield ratio). Given that nitrogen typically limits photosynthetic production, if carbon and nitrogen are upwelled in the same Redfield proportion, unless other dynamics cause an excess of nitrogen, photosynthesis might only assimilate upwelled carbon but not enough to account for all the additional anthropogenic carbon.

However calcifying organisms like pteropods, coccolithophores and foraminifera export greater proportions of inorganic carbon because their sinking calcium carbonate shells lack nitrogen. This can create an excess of nitrogen relative to upwelled carbon in surface waters. Second diazotrophs are organisms that convert atmospheric nitrogen into biologically useful forms. Free-living diazotrophs like the cyanobacterium Trichodesmium, can be so abundant their blooms are readily observed. (The blooms of one species are primarily responsible for the coloration of the Red Sea.) Some diazotrophs form symbiotic relationships with diatoms and coral. So diazotrophs can cause an excess of nitrogen that allows photosynthesis to assimilate both upwelled carbon and anthropogenic carbon. Furthermore as discussed in Mackey 2015 (and references therein) “To date, almost all studies suggest that N₂ fixation will increase in response to enhanced CO2.”

With all things considered, the evidence suggests NOAA scientists have an upside down characterization of the ocean’s “steady state.” There is no rigid rate of primary production and export that prevents assimilating anthropogenic carbon and pumping it to depth. On the contrary the combined dynamics of nitrogen fixation and the biological pump, suggest the upper layers of the ocean have likely maintained a pH homeostasis, or a pH steady state, at least since pre-industrial times. Increases in atmospheric CO2, whether from natural upwelling or from anthropogenic sources, are most likely assimilated quickly and exported to ocean depths where they are safely sequestered for centuries and millennia. As also discussed in the article How Gaia and Coral Reefs Regulate Ocean pH, claims that the upper ocean has acidified since preindustrial times are not measurements, but merely results from modeling a “dead” ocean and ignoring critical biological processes.

Jim Steele is author of Landscapes & Cycles: An Environmentalist’s Journey to Climate Skepticism