Pacific Upwelling Is Dominant Driver of Tropospheric CO2 Growth
In my previous article titled From Lambing Fields to Deep Oceans and Plate Tectonics: A New Take on CO2 and Climate Reality, I introduced the idea of the The Meridional Overturning Mediated Carbon Cycle (MOMCC) as an alternative to the mainstream anthropogenic centric carbon cycle model
If you have not read this Part I in a III part series, I strongly advise that you do so before proceeding this article (Part II).
The two most important take-away points that were introduced in Part I are as follows:
- Deep water, after it forms in the sup-polar North Atlantic Ocean, takes over 2,300 years to transit the ocean depths, before it arrives at and upwells back to the surface in the eastern Pacific Cold Tongue region.
- The deep ocean conveyor or meridional overturning circulation (MOC), follows the path of increasing mid-ocean ridge seismic and tectonic spreading rate activity. This suggests that enhanced mid-ocean ridge hydrothermal flux adds both carbon (e.g., CO2, CH4) and nutrients to deep water over the thousands of years that it transits the abyssal plain in route to the major upwelling zones globally.
Here in Part II, I will produce a body of evidence that clearly demonstrates that the tropospheric CO2 growth rate (TGR) is dominated by the rate by which deep water upwells and off-gases in the Pacific Cold Tongue region.
This of course, stands in complete opposition to the idea that the TGR is increasing solely because of human influence and principally due to our use of hydrocarbons.
The cover image and Figure 1 are taken from the 2017 publication by A. Chatterjee et al, which made use of the recently launched OCO-2 satellite that was used to focus on tropospheric CO2 concentrations over the Cold Tongue and Nino3.4 region (5°S–5°N, 170°–120°W) of the eastern tropical Pacific Ocean.
The timing of this study coincided with the powerful March 2015 to May 2016 El Nino Modoki event.
Starting with the cover image – the upper image (A) represents normal (ENSO neutral) conditions, where Walker Circulation and the easterlies (Trade Winds) are blowing strong along the near equatorial latitude. Note the steep thermocline, the vigorous upwelling, higher CO2 off-gassing and lower sea surface temperatures (SST) in the Nino3.4 region (i.e., less red).
The lower image (B) represents an El Nino condition, which results in the very opposite change in these aforementioned parameters.
Likewise, note that Chatterjee et al acknowledges that deep water is high in carbonic acid (i.e., dissolved CO2).
As Figure 1 contains a tremendous amount of information, I will discuss in bullet point format, each of the four time series – note the dashed vertical line indicating the transition point from neutral conditions to an El Nino state.
- (A) represents the Nino 3.4 indicator or index versus the Southern Oscillation Index (SOI). Note that the definition of an emergent El Nino state in the Nino 3.4 region is 5 consecutive 3-month rolling averaged periods (e.g., JFM, FMA, MAM, AMJ, MJJ) where the Niño 3.4 SST anomaly averages ≥0.5°C above normal. Note that the SOI reflects the pressure difference between Tahiti (eastern Pacific) and Darwin (western Pacific), indicating air mass shifts that drive El Niño (negative) or La Niña (positive) conditions. When there is intense downwelling of atmospheric circulation over the eastern Pacific, the pressure is higher than it is over the western Pacific – downwelling is high pressure and upwelling (convection) is low pressure – the Trade Winds flow from high to low pressure under neutral neutral condition or it intensifies during La Ninas.
- (B) represents the deviation from average of the tropospheric CO2 in parts per million (ppm) over the Cold Tongue and Nino3.4 region – measured using the OCO-2 satellite. This time series shows that initially, tropospheric CO2 drops below average as the Nino3.4 region begins to transition into an El Nino state. As the western Pacific region experiences more intense sunshine and reduced precipitation when the Trade Winds stall across the Pacific tropics and the direction shifts westward, smoke from forest fires in SE Asia and Indonesia begins to move eastward – thus tropospheric CO2 concentrations rise when the peak SST state in the Nino3.4 region occurs.
- (C) represents the fugacity (△pCO2 or fCO2) of CO2 saturated deep water beneath the Nino3.4 region. Philosophically, fugacity can be thought as the pressure difference between carbonic acid enriched deep water and CO2 in the troposphere. Much like the difference in pressure between a balloon and the outside air. When fugacity of deep water is positive, it is analogous to a pressurized balloon. Negative fugacity would be equivalent to a balloon with negative air pressure inside. Like cracking open a can of beer, saturated carbonic acid deep water spontaneously off-gases CO2 when it upwells to the surface. When an El Nino state begins to emerge in the Nino3.4 region, upwelling subsides and CO2 off-gassing decreases. By the time the peak SST condition occurs in the Nino3.4 region, deep water fugacity drops to near zero. Note that Chatterjee et al uses TAO Mooring (Tropical Atmosphere Ocean Array) data to estimate △pCO2. TAO mooring is a deep-ocean buoy array anchored in the equatorial Pacific that continuously measures key ocean-atmosphere variables to monitor El Niño–Southern Oscillation (ENSO). TAO buoys provide real-time data for the Niño 3.4 index, detecting El Niño (warm anomalies) and La Niña (cool anomalies). They capture upwelling changes, thermocline depth, and CO₂ flux.
- (D) represents measured carbon monoxide (CO)concentrations over the Nino3.4 region relative to the onset of an El Nino condition developing. As shown, CO levels spike approximately 4 months after an El Nino state ensues and as stated earlier, this coincides with the development of extensive forest fires in Indonesia and SE Asia as precipitation plummets following the reduction of the Trade Winds that act to maintain low pressure air masses over the Western Pacific Warm Pool region.
Figure 1. Changes in environmental conditions over the Nino3.4 and Cold Tongue regions in the eastern Pacific region.
The next data set that I bring to bear to demonstrate that Nino3.4 and Cold Tongue region show elevated CO2 fugacity’s (fCO2) and CO2 emission rates during non-El Nino conditions and reduced states under El Nino states is shown in Figure 2.
Figure 2 shows TAO Mooring data for fCO2 and CO2 flux and it demonstrates that during El Ninos, CO2 flux drops and drops so far, that some areas actually become net absorbers of tropospheric CO2.
Figure 2. Fugacity (fCO2) and CO2 flux data from the eastern Pacific region under both neutral (top) and El Nino (bottom) conditions.
The next data set that demonstrates that SSTs, fCO2 and CO2 flux plummet when the Trade Winds (aka Walker Circulation) decelerate, is shown in Figure 3. Figure 3 represents the linchpin of the 2006 R.A. Freely et al study that examines the decadal variability of CO2 emissions from the eastern Pacific region and it clearly shows that when the easterlies decelerate, SSTs rise, while both fCO2 and CO2 flux decrease.
The y-axis for each of the four time series, represents the longitude extending from the central (170E) to the eastern (100W) Pacific and each graph has its distinct calibration scale shown on the right – the units for each are listed in the top left hand corner. Refer to the embedded notes in Figure 3 for a deeper perspective on the data shown in each four time series.
Note how even minor deceleration of the Trade Winds speed give rise to these changes.
Figure 3. Seminal study showing how eastern Pacific sea surface temperatures (SST), deep water fugacity (fCO2) and CO2 flux vary as a function of the speed of the easterlies over the eastern Pacific region.
You may be asking, how do these eastern Pacific focused parameters relate to more global average changes in tropospheric CO2 concentration? You are right in asking such as question
First, let me emphasize that there is no such thing as a globally averaged tropospheric CO2 concentration time series, as individual stations are highly synchronized on an inter-annual basis – one station is taken as representative of a global average.
To extrapolate beyond the eastern Pacific region, and to show that these relations hold true, we must use logic and station data in either hemisphere far from the eastern Pacific to infer.
The first such inference that I introduce is shown in Figure 4, which compares the seasonally detrended (12-month rolling average) month-over month rate of change (aka first derivative) of the South Pole tropospheric CO2 concentration time series against the Nino3.4 SST index.
I have introduced the red and green rectangles to aid the eye.
Figure 4. Nino3.4 Index versus the South Pole tropospheric CO2 growth rate (TGR).
The red rectangles capture events where the Nino3.4 slope is at a local maximum positive value, which likewise corresponds with a local minimum in the TGR of CO2 concentration.
The green rectangles capture events where the Nino3.4 slope is at a local maximum negative value, which likewise corresponds to a local maximum in the TGR of CO2 concentration.
Note that there is over 10,000 km between these two regions.
This data shows that even as far away as the South Pole, the TGR of CO2 is seen to achieve a maximum state when La Nina prevail and it approaches a local minimum during El Ninos. This is exactly what Chatterjee et al found for tropospheric CO2 concentrations directly over the Nino3.4 and Cold Tongue Region of the eastern Pacific.
This suggests that tropospheric CO2 emission from the eastern Pacific are transported quickly to the South Pole.
Now that we have established that there is close agreement between the seasonally detrended rates of change in tropospheric CO2 concentration over both the eastern Pacific and the South Pole, it is important to examine the seasonal variation of tropospheric CO2 concentrations between stations in both the Northern and Southern Hemisphere.
Figure 5 shows the monthly variation of tropospheric CO2 concentrations at both Mauna Loa Hawaii and at the South Pole. These plots have not been treated using a 12 month rolling average and thus retain their Seasonal Cycle influence.
Note how both time series are out of phase.
The reason these two stations show that their annual peaks and troughs in tropospheric CO2 concentration are out phase, is because when winter is occurring in one hemisphere, the other hemisphere is experiencing summer.
During the winter season, CO2 concentrations rise and during the summer they plummet. The difference in amplitude between either hemisphere is commonly believed to be due to the difference in ocean versus land area between the Northern versus the Southern hemispheres.
It can be said that these monthly averaged seasonal time series represent the respiration rate of the Biosphere and shows the first order influence of orbital mechanics on the high frequency pace of change in tropospheric CO2 concentration.
The respiration rate of the Biosphere is ultimately controlled by the relative rates of photosynthesis and organic material decay across the oceans and over land.
Figure 5. Monthly averaged tropospheric CO2 concentrations at the South Pole versus the Mauna Loa monitoring stations – seasonally undetrended.
It is important to emphasize that Figure 5 represents what we call intra-annual changes in tropospheric CO2 concentrations. The application of a simple 12 month rolling average low pass filter to these time series effectively dampens the Season Cycle’s signature and the use of a month-over-month difference (aka first derivative) is by definition, how we arrive at the inter-annual CO2 TGR time series.
The inter-annual CO2 TGR time series for both the Mauna Loa and the South Pole stations are shown in Figure 6, where the units for both have been modified by application of a factor to convert changes in concentration to changes in mass (i.e., delta ppm to delta tonnes per month).
The key take-away points from Figure 6 are as follows:
- Peak rates of change coincide with the emergence of intense La Nina (cooling) states in the eastern Pacific region.
- Minimum rates of change coincide with the emergence of intense El Ninos (warming) in the eastern Pacific region.
- There is excellent temporal alignment in the TGR measured at stations in either hemisphere.
- The TGR regularly showed negative rates of change during the 1960s and 1970s, and nearly reached zero in the 1990s – especially true for the Northern Hemisphere where most of human use of hydrocarbons are localized.
If the growth in tropospheric CO2 is due to human use of hydrocarbons, we would not expect to see these features for the following reasons:
- As there is little exchange in atmospheric stream flows across the intertropical convergence zone (ITCZ) and most of humanities’ use of hydrocarbons occurs in the Northern Hemisphere, we would expect the TGR in either hemisphere to be markedly different from each other. The fact that the TGR in either hemisphere is a near mirror image to the other in terms of temporal synchronicity, would suggest that both hemispheres share a common dominating emission source – the eastern Pacific Cold Tongue region. This point is of paramount importance and is virtually ignored.
- Human CO2 emissions were estimated to be 15, 20 and 25 billion tonnes per year in 1965, 1975 and 1995, respectfully, yet Figure 6 shows the Northern Hemisphere’s TGR went negative or was almost zero in those time frames. The fact that carbonic acid enriched deep water upwelling rates approached local minimum values in 1965, 1975 and 1995 during the El Nino states that occurred around these time frames, can not be ignored. This suggests that even when Mother Nature is off-gassing minimal CO2 from the eastern Pacific, her CO2 emissions still still dwarf human CO2 emissions. The fact that the Mauna Loa stations shows the lowest TGR values, can not and should not be over-looked – again the vast majority of human CO2 emissions occur in the Northern Hemisphere.
Figure 6. Tropospheric CO2 growth rates at South Pole and Mauna Loa monitoring stations versus the latitudinal distribution of human CO2 emissions.
To test further the idea that Walker Circulation is first order in setting the pace by which carbonic acid enriched deep water is pulled to the surface and off-gases in the eastern Pacific, I introduce NCEP reanalysis data specific to the Pacific Trade Winds along the narrow track of ocean between 240W – 90W and 5N x 5S (equatorial Pacific).
Note that while this zonal area covers much of the Cold Tongue region, it is not all inclusive, as there are extensive upwelling zones along southwestern coast of United States and Latin America, extending all the way down the coast of South America that fall outside this wind speed zone.
As deep colder water has a higher density than warm surface water, the only way dense water can make its way to the surface is if it is pumped or through geothermal bouyancy enhancement. This is the role of the Pacific easterlies (Trade Winds) and meridional coastal wind fields, which give rise to a surface suction called Ekman Pumping that is proportional to the surface velocity.
As the easterlies move westward along the Pacific equatorial zone, warm surface water is displaced toward Indonesia and carbonic-acid-enriched deep water is pulled to the surface (i.e., cover image).
Figure 7 compares the TGR from Maun Loa and South Pole versus the first derivative of the seasonally detrended velocity of the Pacific trade winds. Recall from first-year calculus that the first derivative of velocity is acceleration. Thus, an acceleration of the easterlies is equivalent to an increase in surface suction or Ekman Pumping rate – conversely for a deceleration of the easterlies.
While there is not a perfect correlation between the two time series, it is clear that La Nina events involve the acceleration of the Pacific easterlies and positive increase in the TGR, which is aligned with the broader literature (e.g., Chatterjee, Freely).
Figure 7. The tropospheric CO2 growth rate (TGR) at Mauna Loa and South Pole versus the first derivative of the Pacific easterlies anomaly.
The next body of evidence that adds credence to the MOMCC and to the idea that the TGR is dominated by changes in Walker Circulation, is isotopic composition of both deep water carbonic acid and CO2 associated with the TGR.
Figure 8 shows modern and paleo-data specific to dissolved inorganic carbon (DIC), pH, and carbon isotopic composition (δCO2) as functions of ocean depth.
First some definitions are in order:
- ∑[δCO2 ] or DIC – represents the total concentration of carbonic acid (H2CO3), bicarbonate (HCO3–), and carbonate (CO32-), which in Figure 8 are shown to increase in concentration with depth in the ocean.
- δCO2 – called the carbon isotopic ratio, which represents the carbon-13 to carbon-12 ratio and where the carbon-13 isotope has one neutron more than carbon-12. By convention, a decreasing δCO2 implies carbon-12 enrichment, whereas an increasing δCO2 implies carbon-13 enrichment.
The mainstream model for the carbon cycle maintains that the declining δCO2 in the atmosphere reflects the combustion rate of carbon-13 depleted hydrocarbons. My model acknowledges that all forms of geological organic and inorganic carbon are depleted in carbon-13.
Figure 8 illustrates data from two studies that show from both the paleorecord and the modern era, DIC (∑[CO2]) in the ocean is higher at depth and the lowest at the surface. Likewise, the paleorecords show that δCO2 and ∑[CO2] are inversely proportional, which implies that carbon-12 is enriched as DIC increases with depth.
Likewise, modern data shows that the pH of sea water decreases with increasing depth, which was discussed extensively in From Lambing Fields to Deep Oceans and Plate Tectonics: A New Take on CO2 and Climate Reality.
Figure 8. Modern and paleo-data showing dissolved inorganic carbon (∑[CO2]), pH, and carbon isotopic δCO2 data versus depth.
When we apply the same data processing techniques used to generate the TGR time series shown in Figure 6, on the δCO2 data from both Mauna Loa and South Pole stations, we find striking evidence that further supports the idea that the TGR reflects the pace by which ancient saturated carbonic acid in deep water upwells and off-gases at the surface in the eastern Pacific.
Figure 9 shows the first derivatives of the seasonally detrended isotopic ratio (dδCO2/dt) at the South Pole and Mauna Loa monitoring stations.
Note that even though the Mauna Loa and South Pole monitoring stations are separated by over 12,000 km, are separated by the ITCZ barrier and most anthropogenic CO2 emissions are in the Northern Hemisphere, their dδCO2/dt are remarkably well synchronized.
Also note that both stations show a maximum negative dδCO2/dt during La Nina events and conversely during El Ninos.
Figure 9. First derivatives of the seasonally detrended isotopic ratio (dδCO2/dt) at the South Pole and Mauna Loa monitoring stations.
Figure 10 shows that when we compare the TGR to dδCO2/dt at the South Pole, we find a clear negative correlation. Similar to the negative correlation between ∑[CO2] and δCO2 versus ocean depth that is seen in Figure 8.
The fact that dδCO2/dt is most negative when the TGR reaches a local maximum during intense La Nina conditions, confirms that La Nina conditions accelerate Ekman Pumping of carbon-12-enriched carbonic acid in deep water to the surface, where it off-gases as CO2, causing δCO2 to trend negatively.
Figure 10. First derivative of the seasonally detrended CO2 isotopic ratio (dδCO2/dt) and the tropospheric CO2 growth rate (TGR) at the South Pole.
At this late stage in Part II, I introduce the classical illustration of the counter rotating Hadley Cells (Figure 11), which converge along the Intertropical Convergence Zone (ITCZ). It is the collision of the Hadley Cells in either hemisphere, which minimizes stream flow across the ITCZ from one hemisphere to another.
While not shown in Figure 11, the Trade Winds, Nino3.4 and eastern Pacific Cold Tongue reside along the ITCZ.
This illustration is introduced at the conclusion to Part II, to help demonstrate how the TGR and the first derivative of the seasonally detrended CO2 isotopic ratio (dδCO2/dt) in either hemisphere are temporally synchronized.
If the common area source of tropospheric CO2 growth is the Cold Tongue region in the eastern Pacific, the low carbon-13 containing CO2 emitted from this large upwelling zone, will quickly be dispersed into higher latitudes of either hemisphere by the Hadley Cells.
Two things equal to the same thing are equal to each other.
Figure. 11 Illustration of the Hadley Cell Convergence Zone or Firebox.
In closing and after reading both Part I & II, you now have a better understanding for why the MOMCC in Figure 12 shows carbon flux to be centered along the ITCZ and coupled to meridional overturning of both the oceans and atmosphere.
While it is popular to imply that the gradual enrichment of tropospheric CO2 by carbon-12 isotopes is the smoking gun proving that human use of hydrocarbons is the source of said carbon-12 isotopes, this hypothesis ignores the evidence that I have outlined.
Deep water has been enriched in carbon-12 for eons and the fact that it takes thousands of years for deep water to transit the abyssal plains of the global oceans, there is plenty of time for further carbon-12 enrichment along mid-ocean ridge spreading zones and through a myriad of other geochemical exchange processes with the Lithosphere.
Figure 12. Meridional Overturning Mediation Carbon Cycle – Dr. Joseph Fournier
Note that I have accepted a senior fellowship role with the Frontier Center for Public Policy and while I promise to continue to regularly write, the frequency going forward will not be the weekly that I was averaging prior to October.
See more here substack.com