Wednesday, April 5, 2017

Falling Sea Level: The Critical Factor in 2016 Great Barrier Reef Bleaching!

It is puzzling why the recent 2017 publication in Nature, Global Warming And Recurrent Mass Bleaching Of Corals by Hughes et al. ignored the most critical factor affecting the 2016 severe bleaching along the northern Great Barrier Reef – the regional fall in sea level amplified by El Niño. Instead Hughes 2017 suggested the extensive bleaching was due to increased water temperatures induced by CO2 warming. 

Reef at Low Tide Around Lizard Island Great Barrier Reef

In contrast in Coral Mortality Induced by the 2015–2016 El-Niño in Indonesia: The Effect Of Rapid Sea Level Fall by Ampou 2017, Indonesian biologists had reported that a drop in sea level had bleached the upper 15 cm of the reefs before temperatures had reached NOAA’s Coral Reef Watch bleaching thresholds. As discussed by Ampou 2017, the drop in sea level had likely been experienced throughout much of the Coral Triangle including the northern Great Barrier Reef (GBR), and then accelerated during the El Niño. They speculated sea level fall also contributed to the bleaching during the 1998 El Niño. Consistent with the effects of sea level fall, other researchers reported bleaching in the GBR was greatest near the surface then declined rapidly with depth. Indeed if falling sea level was the main diver in 2016’s reef mortalities, and this can be tested, then most catastrophic assertions made by Hughes 2017 would be invalid.

Indeed the Great Barrier Reef had also experienced falling sea levels similar to those experienced by Indonesian reefs.  Visitors to Lizard Island had reported more extreme low tides and more exposed reefs as revealed in the photograph above, which is consistent with the extremely high mortality in the Lizard Island region during the 2016 El Niño. Of course reefs are often exposed to the air at low tide, but manage to survive if the exposure is short or during the night. However as seen in tide gauge data from Cairns just south of Lizard Island, since 2010 the average low tide had dropped by ~10 to 15 cm.  After previous decades of increasing sea level had permitted vertical coral growth and colonization of newly submerged coastline, that new growth was now being left high and dry during low tide. As a result shallow coral were increasingly vulnerable to deadly desiccation during more extreme sea level drops when warm waters slosh toward the Americas during an El Niño. 

Furthermore, an El Niño in the Coral Triangle not only causes a sudden sea level fall, but it also generates a drier high-pressure system with clear skies, so that this region is exposed to more intense solar irradiance. In addition, El Niño conditions reduce regional winds that drive reef-flushing currents and produce greater wave washing that could minimize desiccation during extreme low tides. And as one would predict, these conditions were exactly what were observed during El Niño 2016 around Lizard Island and throughout the northern GBR.

Aerial surveys, on which Hughes 2017 based their analyses, cannot discriminate between the various causes of bleaching. To determine the cause of coral mortality, careful examination of bleached coral by divers is required to distinguish whether bleached coral were the result of storms, crown-of-thorns attacks, disease, aerial exposure during low tides, or anomalously warmer ocean waters. Crown-of-thorns leave diagnostic gnawing marks, while storms produce anomalous rubble. Furthermore aerial surveys only measure the aerial extent of bleaching, but cannot determine the depth to which most bleaching was restricted due to sea level fall. To distinguish bleaching and mortality caused by low tide exposure, divers must measure the extent of tissue mortality and compare it with changes in sea level. For example, the Indonesian researchers found the extent of dead coral tissue was mostly relegated to the upper 15 cm of coral, which correlated with the degree of increased aerial exposure by recent low tides. Unfortunately Hughes et al never carried out, or never reported, such critical measurements.

However a before-and-after photograph presented in Hughes 2017 suggested the severe GBR bleaching they attributed to global warming primarily happened between February and late April. Their aerial surveys occurred between March 22 and April 17, 2016. And consistent with low tide bleaching, that is exactly the time frame that tide tables reveal reefs experienced two bouts of extreme low tides coinciding with the heat of the afternoon (March 7-11 & April 5-10). And such a combination of sun and low tide are known to be deadly.

A study of a September 2005 bleaching event on Pelorous and Orpheus Islands in the central GBR by Anthony 2007, Coral Mortality Following Extreme Low Tides And High Solar Radiation, had reported extreme deadly effects when extreme low tides coincided with high solar irradiance periods around midday. As in Indonesia, they also reported bleaching and mortality had occurred despite water temperatures that were “significantly lower than the threshold temperature for coral bleaching in this region (Berkelmans 2002), and therefore unlikely to represent a significant stress factor.” Along the reef crests and flats, “40 and 75% of colonies in the major coral taxa were either bleached or suffered partial mortality. In contrast, corals at wave exposed sites were largely unaffected (<1% of the corals were bleached), as periodic washing of any exposed coral by waves prevented desiccation. Surveys along a 1–9 m depth gradient indicated that high coral mortality was confined to the tidal zone.” [Emphasis mine]

The fortuitous timing of Ampou’s coral habitat mapping from 2014 to 2016 in Bunaken National Park (located at the northwest tip of Sulawesi, Indonesia) allowed researchers to estimate the time of coral mortality relative to sea level and temperature changes. Ampou reported that in “September 2015, altimetry data show that sea level was at its lowest in the past 12 years, affecting corals living in the bathymetric range exposed to unusual emersion. By March 2016, Bunaken Island (North Sulawesi) displayed up to 85% mortality on reef flats” and that almost “all reef flats showed evidence of mortality, representing 30% of Bunaken reefs.” Based on the timing of reef deaths and changes in temperature they concluded, “the wide mortality we observed can not be simply explained by ocean warming due to El Niño.”  They concluded, “The clear link between mortality and sea level fall, also calls for a refinement of the hierarchy of El Niño impacts and their consequences on coral reefs.”

From the illustrations (below) of a generalized topography of a fringing or barrier reef, we can predict the effects of low sea level by examining where bleaching and mortality would occur within the whole reef system. Coral occupying the reef crests are most sensitive to drops in sea level and desiccation because they are first to be exposed to dangerous periods of aerial exposure and last to re-submerge. The inner reef flats are vulnerable to lower sea levels, as those shallow waters are more readily exposed at low tide because the reef crest prevents ocean waters from flooding the flats. If reefs flats are not exposed, the shallow waters that remain can heat up dangerously fast. Accordingly Anthony 2007 found 40 to 75%, and Ampou 2017 found 85% of the reef flats had bleached. In contrast coral in the fore reefs are the least vulnerable to desiccation and higher temperatures due to direct contact with the ocean, upwelling and wave washing. Accordingly Anthony 2007 reported <1% bleaching in the fore reefs.


Coral mortality due to a drop in sea level leaves other diagnostic telltale signs such as micro-atoll formation. As illustrated below in Fig. 4 from Goodwin 2008, during neap low tides (MLWN) sea water can still pass over the reef crest and flush the inner reef with relatively cooler outer ocean water. However during the low spring tides (MLWS), the reef crest is exposed and ocean water is prevented from reaching the reef flats. As mean sea level falls (MSL), coral on the crest and flats are increasingly exposed to the air for longer periods, and the upper layer of coral that had previously kept up with decades of rising sea level, are now exposed to increasing periods of desiccation and higher mortality.

There are over 43 species in the coral triangle that can be characterized as “keep-up” coral whose growth rates are much greater than average 20th century sea level rise. However their vertical growth is limited by the average low water level (HLC-Height of Living Coral in Fig. 4). Average low water level is calculated as the mean water level between low neap tides and lower low spring tides. (Due to the linear alignment of the sun, earth and moon and the resulting stronger gravitational pull during a full and new moon, spring tides result in both the highest high tides and lowest low tides. In contrast neap tides exert the least gravitation pull.  Spring tides typically happen twice a month, but usually no more than once a month will spring low tides coincide with the heat of the midday sun.)

When growing in deeper waters, a keep-up species like mounding Porites spp. grow at rates of 5 to 25 mm per year and form dome shaped colonies. However due to increased aerial exposure when growth reaches the surface, or due to exposure from sea level fall, the upper most surface dies from high air temperatures, higher UV damage and desiccation. This results in a flat-topped colony leading to the classic “micro-atoll” shape, with dead coral in the center surrounded by a ring of live coral, as exemplified by a Kiribati micro-atoll in the photograph below.


Micro-atoll patterns have been crucial for reconstructing past fluctuations in sea level on decadal to millennial timeframes. As Ampou 2017 observed in Bunaken NP, mortality due to a drop in sea level was mostly restricted to the upper 15 cm of coral, which leads to the formation of micro-atolls.  So before simply assuming climate-change-warming has induced mortality, micro-atoll formation and other associated patterns indicative of sea level change must be examined. A short discussion on how sea level changes can shape micro-atolls can be read here.

Due to its regional sensitivity to the sea level change that accompanies an El Niño, the northern Great Barrier Reef has an abundance of fossil micro-atolls that have allowed researchers to estimate El Niño activity and fluctuating sea levels over the past 4000 years. They estimated 4000 years ago low water neap tides were at least 0.7 meters higher than they are at present. Studies of micro-atolls in the Cook Islands further to the east in the southern Pacific, suggest that by 1000 AD during the Medieval Warm Period, average sea level had fallen, but remained about 0.45 meters higher than today. During the Little Ice Age sea level fell to 0.2 meters below current levels during the late 1700s and early 1800s, before recovering throughout the 1900s.

Hughes 2017 wanted to emphasize GBR bleaching as a “global-scale event” in keeping with his greenhouse gas/global warming attribution, but bleaching and mortality was patchy on both local and regional scales. And although Hughes presented their analyses as “a fundamental shift away from viewing bleaching events as individual disturbances to reefs,” the unusually high mortality around Lizard Island demands a closer examination of individual reef disturbances. The lack of mortality in 2016 across the southern and Central GBR, was explained as a result of the cooling effects of tropical storm Winston, but that does not explain why individual reefs in those regions have not bleached at all, while others bleached only once, and still others bleached twice or three times since 1998. Hughes’ shift away from examining what factors affected individual reefs will most likely obscure the most critical factors and yield false attributions.

Hughes reported the various proportions of areal bleaching as degrees of severity. But that frightened many in the public who confused bleaching with mortality, leading some misguided souls to blog the GBR was dead.  However bleaching without mortality is not a worrisome event no matter how extensive. Rates of mortality and recovery are more important indices of reef health. As discussed in the article The Coral Bleaching Debate: Is Bleaching the Legacy of a Marvelous Adaptation Mechanism or A Prelude to Extirpation?, all coral retain greater densities of symbiotic algae (symbionts) in the winter but reduce that density in the summer, which often leads to minor seasonal bleaching episodes that are usually temporary. Under those circumstances coral typically return to normal within weeks or months. Furthermore by ejecting their current symbionts, coral can acquire new symbionts that can promote greater resilience to changing environmental conditions. Although symbiont shifting and shuffling promotes adaptation to shifting ocean temperatures, symbiont shuffling cannot protect against extreme low tide desiccation, and dead desiccated coral can no longer adapt. Humans have little control over El Niños or low tides.

Hughes also contradicted past studies to mistakenly suggest that recurring bleaching in a given reef is evidence that corals are not adapting or acclimating. However bleaching happens for many reasons. Symbiont shuffling to better adapt to warmer waters does not guarantee adaptation to lower sea levels, cyclones or changes in salinity. Coral reefs deal with changing sea levels with rapid growth to keep-up as sea level rises, and then dying back when sea level falls. Decadal swings in regional sea level will likely cause decadal swings in bleaching and are not evidence of coral fragility.

Hughes 2017 modeled the 2016 GBR bleaching event as a function of surface ocean temperatures that surpass bleaching thresholds, although reefs will bleach below that threshold and will fail to bleach despite temperatures above that threshold. Despite the fact El Niños are well known to cause rapid sea level fall along the GBR, Hughes’ model never accounted for falling sea level. Nor did they account for past observations that falling sea levels induced bleaching when temperatures were below bleaching thresholds. More disturbing because sea level fall caused bleaching in various reefs, with some experiencing good water quality and others poor quality, Hughes asserted there was “no support for the hypothesis that good water quality confers resistance to bleaching.” However this contradicts an abundance of regional studies attributing increased coral disease and bleaching to high nutrient loading.

Woolridge 2013 have argued that coral eject their algal symbionts and bleach when temperature, light and nutrients increase to a level that accelerates the symbionts growth. Increased growth consequently reduces the amount of energy transferred to the coral, resulting in ejection of the slacking symbiont. Because increased nutrient loads can promote increased symbiont growth at relatively lower temperatures, higher nutrient loads can promote bleaching at lower temperatures.

Furthermore while coral’s symbiotic relationships allow them to recycle limited nutrients and out compete seaweeds, higher nutrient loads enable greater seaweed growth, which reduces corals’ competitive advantage. Furthermore seaweeds have been shown to harbor allelopathic chemicals that inhibit coral growth, as well as serving as reservoirs for bacteria that cause coral diseases. Higher nutrient loads induce more dissolved organic carbon that bacteria feed upon, allowing disease-causing bacteria to rapidly multiply. Higher nutrient loads also increase the survival of crown-of-thorns larvae, which then increases coral depredation and bleaching.

In a 2013 experimental study, Chronic Nutrient Enrichment Increases Prevalence And Severity Of Coral Disease And Bleaching, Vega-Thurber reported that higher nutrient loads caused a “twofold increase in both the prevalence and severity of disease compared with corals in unenriched control plots” as well as a “3.5-fold increase in bleaching frequency relative to control corals.”

Although Hughes 2017 suggests the pattern of recurring bleaching is simply a function of temperature and global warming, as illustrated in Hughes’ Figure “e” below, recurring bleaching is not a global phenomenon. (Black dots represent reefs that bleached during all 3 surveys: 1998, 2002, 2016; light gray represents reefs that bleached only once, and dark gray reefs bleached twice.) . In most cases the degree of recurring bleaching does not predict the recurrence of bleaching in nearby reefs despite similar ocean temperatures. Although an El Niño generates widespread bleaching, bleaching is still a regional issue affecting individual reefs differently. During an El Niño sea level rises in the eastern Pacific and falls in the western Pacific. Recurring bleaching in the Far North and Southern regions of the GBR are uncommon, while recurring GBR bleaching has been frequent between Cookstown and Townsville where temperatures have been quite variable. And in accord with prior research, the region between Cookstown and Townsville has suffered from lower water quality and higher nutrients loads, causing more frequent bleaching and greater crown-of-thorns attacks.

Reefs visited during 3 surveys and recurring bleaching

After perusing Hughes 2017, it was clear they had been led to incorrectly embrace the prevailing bias of  CO2-induced catastrophic bleaching because they failed to address the fall in sea level before and during the 2016 El Niño, and likewise they failed to address how weather created by El Niños promotes clear skies and increased solar heating. To add insult to injury, because sea level drops bleached reefs in both good water quality and bad, and bleaches reefs in both protected preserves and unprotected, Hughes 2017 presented a statistical argument that disparaged any significant value of ongoing conservation efforts to minimize bleaching by reducing nutrient loading and by protecting reefs from overfishing. By belittling or ignoring most critical factors affecting coral bleaching other than temperature, Hughes suggested our only recourse to protect reefs “ultimately requires urgent and rapid action to reduce global warming.”

And because such an apocryphal analysis was published in Nature and will undoubtedly mislead coral conservation policies,

I wept.

Friday, March 3, 2017

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

 Are the Oceans’ Upper Layers Really Acidifying?

Bad models, not measurements, have suggested ocean acidification in the upper layers of the oceans. 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. 

Sea Butterfly  Limacina helicina

(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.

Daily CO2 concentrations in surface waters off Newport Oregon from Evans 2011

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.

Estimation of sequestration depths and proportion of carbon reaching sequestraton

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.

Tuesday, February 28, 2017

How NOAA and Bad Modeling Invented an “Ocean Acidification” Icon: Part 1 - Sea Butterflies

If you google “ocean acidification,” the first 3 websites presented according to “Google’s truth rankings” are: 1) Wikipedia, 2) NOAA’s PMEL site featuring the graphic cartoon shown below with a dissolving pteropod shell (a sea butterfly) as the icon of ocean acidification, and 3) the Smithsonian’s Ocean Portal site similarly featuring a dissolving sea butterfly shell. However NOAA’s illustration incorrectly implies shells are dissolving near the surface due to invading anthropogenic atmospheric CO2. As will be shown, the depiction would be far more accurate if it was turned upside down, so that the downward arrows point upwards to illustrate shell dissolution happens when old carbon stored at depth is upwelled to the surface. Furthermore the horizontal depiction of extreme dissolution illustrated by their intact (green) sea butterfly shell dissolving into an extremely shriveled shell (red), rarely if ever happens in the ocean’s upper layers. Surface waters are supersaturated regards calcium carbonates. Although upwelling causes some near surface dissolution, dead sea-butterfly shells only experience such extreme dissolution when they sink to depths containing ancient corrosive waters.

NOAA's Upside Down Cartoon of Sea Butterfly SHell Dissolution

As for most organisms, pteropod populations fluctuate over the short term. But research finds no significant long-term trends in pteropod abundance. Nonetheless NOAA’s Nina Bednarsek has been preparing a preliminary report arguing sea butterflies should be listed as endangered and NOAA’s cartoon appears to be an attempt to gain support for her claims. To warrant endangered status, Bednarsek presents a hypothesis that increasing CO2 has reduced critical pteropod habitat by raising the depth of calcium carbonate saturation horizons.  The threshold above which high concentrations of carbonate ions (CO32-) promote abiotic calcium carbonate precipitation but below which favors dissolution, is referred to as the saturation horizon. The horizon is quantified as 1, and higher numbers characterize supersaturated water that favor calcification. As seen below in Fig 10 from Jiang 2016, most of the globe’s surface oceans are supersaturated.

Global calicum carbonate supersaturation from Jiang 2016

Bednarsek assumes anthropogenic carbon is mostly accumulating near the surface based on modeling results. However as detailed in Part 2, all ocean acidification models are deeply flawed based on an incorrect assumption that CO2 enters the ocean and is then transported like an inert tracer. But CO2 is not inert! When CO2 first invades sunlit surface waters, it indeed dissolves into 3 forms of inorganic carbon (DIC) and lowers pH (DIC is discussed in How Gaia and Coral Reefs Regulate Ocean pH). But in contrast to those models, DIC is rapidly assimilated into particulate organic carbon via photosynthesis, which raises pH.  Particulate organic carbon (alive or dead) is heavy, and if not consumed and recycled, it sinks. For millions of years, this process created and maintained a DIC/pH gradient with high pH/low DIC near the surface and low pH/higher DIC at depth.

Gravity drives the biological pump and removes a significant proportion of organic carbon (assimilated from both natural and anthropogenic carbon). That carbon is transported to depths where it can be harmlessly sequestered for hundreds to thousands of years. However NOAA’s models fail to account for the biological pump, based on the narrow belief that carbon storage is strictly “a chemical and physical response to rising atmospheric CO2” (Sabine 2010). In contrast to Bednarsek’s anthropogenic hypothesis, an increase in the assimilation of CO2 and an efficient biological pump can prevent a decrease in surface pH and calcium carbonate saturation. In fact experiments show CO2 is often a limiting nutrient. Mesocosm experiments found that when atmospheric CO2 was increased, primary production by plankton community consumed 39% more DIC. When primary production increases, more carbon is shuttled to depth.

Meet the Sea Butterflies



Sea Butterfly  Limacina helicina


Sea butterflies are pteropods, a kind of snail exclusively living in the open ocean. A cubic meter of seawater may contain 50 to several thousand individuals. Unlike their terrestrial relatives that plod along on a slimy “foot”, pteropods transformed their foot into a pair of wings to “fly” through ocean waters.

Pteropods are divided into two main groups: sea butterflies with extremely thin, coiled or cone-shaped shells, and “naked” sea angels that evolved a way to shut off their shell-making genes completely when larvae. Sea butterflies feed by suspending themselves in the water column and extruding a web of mucus that passively catches sinking plankton and other organic particles. In contrast sea angels specialized to aggressively prey on sea butterflies. Abandoning their shell suggests whatever benefits a shell may have provided, those benefits were not critical, but losing the shell increased their maneuverability for the hunt. When encountering a sea butterfly, the bizarre sea angel shoots out tentacles from its head. The tentacles dig into the butterflies’ shells and if properly grasped, the tentacles give the angel leverage to extract the butterfly from its shell opening. Below is a 2-minute video below of a sea angel attacking a sea butterfly (not in English). Fish and whales also feed on sea butterflies, gulping mouthfuls at a time. So overall the butterfly’s shell offers precious little protection from their main predators.

While Bednarsek fears pteropods might not adapt quickly enough to rising atmospheric CO2, pteropod behavior argues they are already well adapted. Sea butterflies prefer to graze in highly productive regions generated by nutrient rich but corrosive upwelled waters. Accordingly upwelled regions are typically key reproductive habitat supporting an abundance of juveniles. Sea butterflies are most abundant in the upper 50 meters of the ocean, grazing on abundant phytoplankton. However depending on the species, the population, and location, most sea butterflies migrate daily to depths of 100 meters or more (sometimes below 500 meters) where pH can drop to around 7.6 and waters become corrosive. Similarly they will migrate to deeper more “acidic” depths to over-winter. And although tropical waters are the most supersaturated and the most unlikely to promote shell dissolution, pteropods are least abundant in those tropical waters. In contrast they are very abundant in marginally supersaturated waters around Antarctica.

All calcifying organisms have a protective organic layer that minimizes sensitivity to any changes in seawater pH and all isolate their calcifying chambers from ambient water conditions. Mollusks like clams, oysters and snails have a protective outer layer of organic tissue called the periostracum. The mollusk periostracum has allowed them to colonize the acidic depths of ocean floors, colonize freshwater lakes and streams where pH falls to truly acidic levels below 6.0, and to colonize the flanks of submerged volcanoes where escaping CO2 naturally lowers the pH between 7.3 and 5.39.

Single cell foraminifera and coccolithophorids have some of the thinnest organic layers that effectively prevent dissolution, and the petite sea butterfly has one of the thinnest mollusk periostraca. How well it protects the sea butterfly has created a debate between Bednarsek and other pteropod researchers. Bednarsek argues the sea butterfly’s thin periostracum, especially in juveniles, offers very little to no protection from low pH water suggesting they are very susceptible to life threatening shell dissolution. In contrast other researchers argue shell dissolution occurs when the periostracum is damaged. During the sea butterflies’ short life, which can be less than a year, they are under constant attacks from predators like sea angels that can damage their periostracum. There also exists a whole range of shell-inhabiting/shell-digesting organisms from bacteria to sponges and worms (aka the epibiont) that drill through a mollusk’s periostracum. Thus researchers argue when the periostracum remains intact, “the shell appears pristine with no sign of dissolution”, even when exposed to undersaturated waters. Only the damaged shells showed dissolution when exposed to undersaturated water. Based on observations, they concluded sea butterflies “are perhaps not as vulnerable to ocean acidification as previously claimed, at least not from direct shell dissolution.”

Dissolution of Sea Butterfly Shell from Bednarsek 2014

Furthermore to counteract shell dissolution in damaged areas, sea butterflies rapidly repair their shells by adding more calcium carbonate to the inside of the shell. Biogenic calcification happens at much greater rates than dissolution, and such rapid repair mechanisms would be expected for an animal seeking low pH upwelled waters to graze.

Ironically Bednarsek’s electron microscope images of corroded sea butterfly shells, provide evidence that supports her detractors. The shell (Figure 2 above) from Bednarsek’s 2014 paper, shows severe dissolution (labeled “b”) on the innermost whorls, similar to her other images (not shown here) showing widespread dissolution on juvenile shells. However despite dissolution during its juvenile stage, the snail clearly survived and continued to grow. The subsequent growth shows very little dissolution (labeled “a”). That suggests two possible scenarios that are not mutually exclusive. The snail’s exposure to corrosive waters was limited to short-term episodic upwelling during its earliest years, and was an insignificant cost of grazing in highly productive waters. Or if corrosive conditions continued, then the more developed periostracum protected the shell from further dissolution. The small area of severe dissolution within the section of unharmed shell (the patch to the right of region b) further supports the argument that dissolution only happens where there is damage. Otherwise, if the periostracum provided little protection, then the whole region would have suffered dissolution not just the isolated patch. Thus life-threatening dissolution is just conjecture, and as Bednarsek later admits, “dissolution-driven mortality in pteropods has not been directly confirmed.”

Catastrophic Dissolution?

Long before the politics of climate wars emerged, Mark Twain quipped, “There is something fascinating about science. One gets such wholesale returns of conjecture out of such a trifling investment of fact.”  Indeed the oft-cited Orr 2005 “ocean acidification” paper is an iconic example illustrating Twain’s observation. Orr et al examined just 14 living sea butterflies (Clio pyramidata, see photograph below), captured in the subarctic Pacific. Orr exposed them for 48 hours to experimentally induced under-saturated water that he predicted would occur in the future around Antarctica due to rising atmospheric CO2. All specimens survived. However scanning electron microscopy revealed etching on their shells.

Although it is highly unlikely such minor etching would be of any consequence, Orr 2005 jumped to hypothesizing that pteropod species “will not be able to adapt quickly enough to live in the undersaturated conditions”. Orr 2005 lamented rising atmospheric CO2 would reduce pteropod habitat and survival, leading to ecosystem collapse in polar regions. He speculated arctic pteropods could be forced southward to warmer waters that were more saturated. Due to the butterfly’s imagined extirpation via dissolution, he predicted sea butterfly predators such as sea angels, fish and whales would all suffer. Such was their wholesale catastrophic conjecture, published in a top journal, based on etching in 14 specimens. 

Observed Dissolution due to Ancient CO2 Enriched Waters

Nina Bednarsek followed with a 2012 paper in which she too attributed shell dissolution of a sea butterfly (Limacina helicina) to increasing anthropogenic CO2. Media outlets promoted her conclusions in articles like Science Daily’s “First Evidence Of Ocean Acidification Affecting Live Marine Creatures In The Southern Ocean. Unfortunately the term “ocean acidification” is indiscriminately used to describe any reduction of pH even if naturally induced by upwelling.

On her cruise through the Scotia Sea (just north of the Antarctic peninsula), Bednarsek collected snails at depths of 200 meters from 6 different stations. Snails at five stations showed no evidence of shell dissolution. All stations exhibited supersaturated surface waters. Her shell dissolution was limited to just one station.

All stations experienced upwelling and Bednarsek acknowledged, “these upwelled waters are approximately 1000 years old”. Thousand-year old water means it had not contacted the surface for 1000 years, and those deep waters had not absorbed any anthropogenic CO2. The water’s corrosiveness was due to a millennium of decomposing sunken organic matter that increasingly released CO2 and lowered pH. Still she conjectured the observed dissolution was due to anthropogenic CO2.

Because winds will mix supersaturated surface waters with corrosive upwelled waters in the upper 200 meters, mixing can neutralize any corrosive effects of upwelled waters. To attribute her observed dissolution to anthropogenic CO2, Bednarsek argued recent invasions of anthropogenic CO2 into the surface water had lowered its surface pH to such an extent, mixing no longer counteracted the low pH of upwelled water. It was a reasonable hypothesis, however there was no evidence to support it.

At the one and only station where snails had experienced dissolution, surface waters were far more supersaturated than at any other station. In contrast to Bednarsek’s narrative, that station’s supersaturated waters should have had the most neutralizing effect. Her anthropogenic attribution was simply not consistent with observations. More parsimoniously either that one station experienced greater upwelling, or there was less wind mixing to deepen the neutralizing effect. It is puzzling why peer-review or the editors at Nature allowed her unsupported anthropogenic conclusion to be published.

In 2014, NOAA News promoted another paper by Bednarsek about dissolving sea butterfly shells with the headlines, “NOAA-Led Researchers Discover Ocean Acidity Is Dissolving Shells Of Tiny Snails Off The U.S. West Coast.”  NOAA’s press release explicitly stated the term “ocean acidification” described the process of ocean water becoming corrosive as a result of absorbing nearly a third of the carbon dioxide released into the atmosphere from human sources.” In contrast, her observed shell dissolution only happened where upwelling was the greatest, along Oregon and northern California.  Along southern California where upwelling was minimal, Bednarsek found no dissolution. If acidification was due to atmospheric CO2, we would expect a more uniform pattern of dissolution. But again Bednarsek set forth a scenario to blame anthropogenic CO2, arguing upwelled waters had been directly “acidified” by anthropogenic CO2.

She speculated the upwelled waters had been near the surface 50 years ago, during which time it equilibrated with the 1960s atmosphere. Those waters then sank to depths of 80 to 200 meters, and were now upwelled to the surface. However the problem with this scenario is the source of upwelled waters along the California Oregon coast can be traced back to the California Undercurrent. The California Undercurrent originates in equatorial regions primarily at depths 100 to 200 meters and flows poleward beneath the equatorward flowing California Current. The undercurrent is supplied with low pH, low oxygen, and high inorganic carbon waters from the eastern tropical Pacific. Due to the accumulation of ancient carbon, the eastern tropical Pacific contains some of the oldest waters on earth and more than any other region on earth ventilates tremendous amounts of CO2 from the ocean into the air.

Studies of the changing characteristics of the California Undercurrent conclude all its “changes are consistent with an increasing influence of Pacific equatorial waters” over the past decades. During a negative Pacific Decadal Oscillation (PDO) or a La Nina, research shows the California Undercurrent acquires increased amounts of those ancient waters and upwelling is stronger. Although those studies did not consider possible anthropogenic contributions, the trends in lower oxygen and lower pH were explained by natural increased mixing of older water masses. That mixing trend could represent decadal variability or the centuries long increasing upwelling trend documented under the Peru Current (detailed in Part 2).

The same corrosive upwelling associated with Bednarsek’s sea butterfly dissolution was also responsible for the corrosive waters that oyster fisherman unwittingly pumped into larval-rearing tanks in 2008-9. (To make matters worse, oyster fishermen pumped water in the early morning when nighttime respiration further acidified the water). And consistent with multidecadal variability, during the previous negative PDO from 1940 to the late 70s, corrosive upwelled waters had similarly reduced survival of larval oysters in those bays. Clearly natural oscillations episodically upwell more nutrient-rich, oxygen-poor corrosive waters. And because oceans contain the greatest concentration of inorganic carbon by far, the question remains has anthropogenic CO2 significantly exacerbated the corrosiveness of natural upwelling? In contrast, NOAA’s Richard Feely believes anthropogenic CO2 is the primary factor. Thus he answers that question with an upside down perspective, stating upwelling “exacerbates [anthropogenic] ocean acidification”. And as detailed in part 2, NOAA's models have an upside down representation of carbon distribution.

The sea butterfly joins the parade of icons like polar bears, penguins, pika, mangroves and Parmesan’s butterflies where the effects of natural climate variability or direct human interference are obscured and falsely promoted as catastrophic climate change.