I. Introduction

As technology continues to improve, it was inevitable that at some point man would decide to try and counteract some of the damage we have inflicted on our environment via more technology. This concept of climate engineering is filled with unique ideas and concepts, one of which was focused upon for this assignment. At the forefront of these studies are concepts for remedying the apparent pending global warming resultant from combustion of fossil fuels. Is it feasible to use technology to alter the world’s climate? We will look at one hypothesis and consider just this.
 

For a gardener, fertilizer would be a common means toward improving plant growth. Imagine a fertilizer so powerful it could increase your garden yield by over 2500%. Sound good? This is what scientists came across when they "fertilized" a patch of the ocean in 1995 with Iron Sulfate. Why would this chemical have had such a profound effect on phytoplankton growth? It turns out that over 20% of the world’s oceans are nutrient rich but iron poor, which is the limiting nutrient for phytoplankton growth. By adding sufficient levels of iron to the surface ocean water, a phytoplankton bloom can be induced, and the effect can be quite dramatic. During the 1995 IronEX II experiment, 450 Kg of iron was spread over a 100 km2 patch of the ocean, producing a phytoplankton bloom which consumed over 2500 tons of carbon dioxide from the surface ocean waters.
 

The possible implications of this study are quite remarkable. Recognizing the vast surface area of the world’s oceans that are primed for fertilization, one can’t help but wonder just how much carbon could be sped through the Earth’s carbon cycle via this method. Reduction of surface ocean water carbon dioxide concentrations would lead to reduced atmospheric concentrations, and perhaps minimize the extent of global warming. The author wanted to explore this possibility to determine the extent of fertilization which would be required to have a significant impact on the global atmospheric CO2 concentration.
 
 

II. Background

Phytoplankton is a major contributor to the global carbon cycle, accounting for about 40% of the natural consumption of CO2 by biomass. This tends to suggest that the livelihood of plankton would play an essential role in management of the global carbon cycle. Iron is an essential element to the plankton’s existence, among many other chemicals. Generally the plankton grow, consuming CO2 through photosynthesis, die,and are broken down by bacteria returning the nutrients they consumed except for a small fraction of carbon that sinks into the deep oceans or dissolves into the ocean waters and is circulated to the deep oceans.
 

The concept of iron limiting the growth of phytoplankton has been around since the 1930’s. Scientists recognized that iron concentrations would be limited in the oceans due to the high oxygen concentrations and limited sources for iron. Natural sources of iron are generally limited to iron laden dust falling out of the atmosphere, rivers and hydrothermal vents in deep ocean ridges. During the 1980’s, extensive research was led by John Martin into the role that iron played in phytoplankton productivity. His clean sampling techniques made it possible to recognize the extreme lack of iron presence in some ocean waters.
 

Through careful experimentation, Martin was able to amply demonstrate that iron was a limiting factor to plankton growth in some waters. These water are high in nitrates, yet lack much plankton growth, known as High Nitrate Low Chlorophyll (HNLC) waters. HNLC waters primarily exist in the Antarctic Ocean and the equatorial Pacific. Since equatorial Pacific waters eventually drift via currents into areas where iron is not limited, the primary waters to focus on for effective results is the Antarctic Ocean which covers approximately 20% of the ocean surface. Due to the extremely low concentrations of iron in these ocean waters, experiments were difficult due to the risk of contamination. Due to the difficulties in accurate sampling, a test was set up by Martin to determine true ocean sensitivities.
 

The first major success in iron fertilization occurred in 1995 with Iron EX II. For this experiment, a 10 by 10Km patch of ocean near the Galapagos Islands was fertilized due to its ideal control characteristics (abundant sunshine, weak currents). This experiment slowly spread 450 Kg of Iron sulfate into the surface ocean waters for 18 days. A plankton bloom rapidly invoked the entire patch, turning the waters brown with plankton. It was estimated from plankton density samples that the plankton consumed nearly 2500 tons of CO2, significantly reducing the concentration of CO2 in the ocean patch from its original value.
 

There is some supporting evidence of the importance of iron concentrations and global mean temperature. Ice cores taken from the Antarctic have revealed that during ice ages the concentration of iron dust aerosols in the atmosphere is much greater than during warmer eras. An additional factor to consider is a biproduct of phytoplankton growth – dimethyl sulfide. This chemical after reactions in the atmosphere is a potent cloud seeding aerosol and could be expected to increase global cloud cover if released in large quantities.
 
 

III. The Model

The model for this study was constructed using Stella v5.1.1 software. The system was designed with eight reservoirs, with primary emphasis on the oceanic processes. A basic relationship between reservoirs is shown in Figure 1. There are two reservoirs for land processes, five for the ocean processes, and the atmosphere. Initial stores of carbon are given in each box in Gt of carbon. The time period for the run was from present to the year 2100.
 

All flows are donor controlled with the exception of the atmosphere to ocean mixed layer interface and the atmosphere to land plant flow. The flows used for these cases are based on equations suggested by Bolin. The ocean CO2 absorption sensitivity is based on the atmospheric concentration in relation to the vapor pressure in the ocean layer. The relation between the atmosphere and the land plants is also sensitized to increases in CO2 given that higher CO2 concentrations enhance the growth of some plants.
 
 

Figure 1: Basic carbon cycle model for this experiment.
Relations indicate direct flow interactions between the
reservoirs shown in the diagram.


 Some other essential flows within the model (not shown in Figure 1) are the contributions of carbon to the atmosphere from the burning of fossil fuels and cement production, and flows affecting the land biota of deforestation and reforestation. The fossil fuel contribution is designed based on a market controlled consumption rate with the peak consumption occurring after the year 2100. Initial anthropogenic production is set at 5.5 Gt Carbon per year. The flow values for carbon due to deforestation and reforestation are held constant at 1.6 and .5 Gt C/yr respectively throughout the model run for consistency. Within this model, reforestation is defined as the re-growth of forests into the areas previously cleared for farming. A diagram showing all of the reservoirs and their relations (a screen capture of the Stella model) can be found here.
 

Recognized weaknesses with this model include the ocean circulation system. The IPCC accepted model of the ocean is an upwelling-diffusion model, not reservoir based. The linear first order donor controlled relationships may not well represent the actual flow rates between these reservoirs. This becomes quite apparent in the results shown later. Additionally, ocean mean temperature is not taken into account by this model which effects the absorption rate of CO2 from the atmosphere by the ocean mixed layer.
 
 

IV. Model Results

After running the model with the above noted parameters, the results were reasonably encouraging. By fertilizing the entire 20% of the world’s oceans that are primed for this growth, ~38 PPM of CO2 was reduced from the control scenario by the year 2100. The fertilization accounted for 76.4 Gt of carbon removed from the atmosphere via this methodology. This is far from the amount required to balance uncontrolled CO2 emissions, but is still a substantial quantity. A graph showing the results for various percentages of fertilization is shown in Figure 2. The response was found to be nearly linear for this case, which seems reasonable since deep ocean circulations act on time scales near 300 to 500 years, so longer runs might be needed to find the true long term effectiveness of this method.
 
 

Figure 2: CO2 in PPM vs. Percent Ocean Fertilized
Results from the model run showing the amount of carbon sequestered
from the atmosphere in PPM CO2 for varying percentages of the ocean
surface fertilized through the year 2100.


 A concern from this run was where exactly would all of the consumed carbon end up. A table of the reservoir contents with and without fertilization can be found here. This provided an obvious weakness in the reservoir with first-order-flow-rate method implemented here. The author would have preferred to find the carbon transported into the sea floor sediment. However, most of the carbon in fact ends up as dissolved organic carbon in the ocean. The rate of upwelling carbon to the mixed ocean layer is probably incorrectly accelerated in this model. This said, the rate of absorption from the atmosphere would likely have been greater were it not for this flow problem. Therefore, a better treatment of upwelling might enhance the amount of carbon which would have been removed from the atmosphere. This would tend to agree with other simulations, which have suggested a CO2 PPM reduction closer to 60 PPM.
 

The system which we might use to accomplish this large scale fertilization would be no small feat. The amount of ocean surface area being dealt with here is quite substantial, with 72 million Km2 to cover. To accomplish this using ships such as the experiment was conducted, it would require over 300,000 ships and 1.62 billion Kg of Iron annually to accomplish this task. This would obviously not be very feasible, but other options exist. Perhaps a better quality method would distribute a sort of floating time release capsule that served to distribute the iron over an area without requiring continual manning.
 
 

V. Implications

What good would the reduction of only 38 PPM do towards reducing global warming? Perhaps not a great deal, but there is a deeper issue to explore here. With the potential future situation where carbon taxes are implemented globally, there is a great deal to be said for the economic feasibility of various means for carbon removal. It turns out that fertilization might be one of the cheapest ways to remove carbon. Could a country engage in fertilization to count against carbon production within their country? Would this be fair when the cost is borne not so much by the country, but by the ecosystems of the Antarctic, a "commons" resource? These types of issues are likely to come to the forefront in coming years.
 

What effects would resolve from long term fertilization? At present, little is known about the possible reactions to the oceanic ecosystem from this type of activity. From observations there appears to be a correlation between algal blooms and reduction in biodiversity. Similar corelations might exist with phytoplankton. What would happen to the ocean chemistry following large scale fertilization? Quite possibly the added biomass would consume the deep ocean oxygen levels and produce methane as it decayed, a potent greenhouse gas. The added surface matter would also reduce the depth to which sunlight would penetrate, likely upsetting deeper algae growth.
 

There are a many number of different issues that would need to be considered before this type of activity should be considered on a large scale. Some of these might include:
 

As previously discussed, this issue deals with the release of dimethyl sulfide from plankton growth. Once in the atmosphere, this chemical oxidizes to form sulphate aerosols. These aerosols are highly hygroscopic and hence could increase cloud cover if introduced in sufficiently large quantities. The increased cloud cover could lend toward shifting the global radiation budget and lead to a cooling trend for Earth’s mean temperature.
  Little is known of the long term effects to the biota in the sea, only some of the short term effects to the plankton have been have been looked at so far. A great deal of research would need to be conducted to determine the potential risks to the world’s ocean biodiversity prior to implementation of this type of system.
  There was also some speculation that altering plankton growth might alter algae growth which in turn might affect deep ocean iron beds, causing an increase in iron concentrations and further amplifying the effects. If this were an unstable scenario, it might reach a scale sufficient to place the Earth back into an ice age.
  There exists the possibility that the carbon would not be effectively removed. The carbon might simply be absorbed temporarily and then re-released some finite time later as dissolved carbon levels increase in the ocean mixed layer and it would be largely returned to the atmosphere.
 


There are many natural issues to consider as well. Some that might have profound implications on the effectiveness of fertilization both from an artificial and a natural sense include:
 

Another potential source for iron concentration changes to these waters might be shifts in hydrothermal activity deep within the ocean. Deep ocean hydrothermal vents pump large quantities of iron rich water into the ocean. If this iron were to be introduced into the portions of the ocean where it is currently deficient, a natural bloom might take place.
  Shifts in the mean oceanic temperature could have significant effects on the atmospheric concentrations of carbon. The ocean atmosphere interface is highly dependent on the surface ocean temperature for its effectiveness at absorbing CO2 from the air. A rise in mean ocean temperature might offset the effectiveness of fertilization.
  A major change in oceanic circulation patterns might act to reduce or increase the surface area of waters primed for fertilization. Areas of cold, downwelling waters in the Antarctic Ocean might change in flow rate or in mixing with other waters that might have higher iron concentrations.
  My reason for including this topic is related to the current source for iron in the open ocean – settling of iron rich dust. If there were significant changes to the areas of the Earth’s deserts such that newer ones may contribute a higher concentration of iron to airborne dust, this might act to naturally fertilize the worlds oceans without any intervention required.
  These and many other factors may play into the success or demise of the Iron Hypothesis strategy. Ideally the research currently underway will be able to answer some of these questions or find other possible solutions based on the information gathered. Based on the scale of the task at hand, it would be ideal to find a way to gain ‘nature’s cooperation’ in any attempt to fertilize on such a larger scale. Whether this is the ethically correct way to go about solving our carbon emission problems is a more important question that policymakers will hopefully be able to answer once the true risks of this methodology are better understood.
 
 

VI. Conclusion

There are still too many unanswered questions about fertilization for large scale implementation to be considered. However, this issue will likely be coming to the forefront in the not-so-distant future as countries become increasingly concerned with cost-benefit analyses for carbon reduction. Iron fertilization is one of the cheapest means of sequestering carbon from the atmosphere that is currently known, running about $5.00 per ton. Many other methods considered, such as seaweed farming would cost upwards of $200.00 per ton.  New research is currently underway to attempt to replicate the results found during IronEx II.
 

Hopefully the adverse effects to the marine ecology would be minimal and the usage would not result in any unknowable feedbacks. With our limited understanding of the ocean system, our best course of action for now would be to continue small scale experimentation and advance our knowledge level of the ocean ecology system.  But for certain, we need to be working now toward establishing policies to limit our future consumption of fossil fuels and seek out alternative sources of energy.
 
 

VII. Resources

Written

Bolin, Bert - Steady State and response Characteristics of a Simple Model of the Carbon Cycle, 1981

Intergovernmental Panel on Climate Change - An Introduction to Simple Climate Models used in the IPCC Second Assessment Report, Feb. 1997

Kasting, James F. - The Carbon Cycle, Climate, and the Long-Term Effects of Fossil Fuel Burning, Consequences, Vol. 4, Num. 1, 1998

Rockwell, Richard C. - From a Carbon Economy to a Mixed Economy: A Global Opportunity, Consequences, Vol. 4, Num. 1, 1998

Schlesinger, William H. – Biogeochemistry : An Analysis of Global Change Sec. Ed., Academic Press, 1997
 

Web-based

Chang, Gloria - Fertilizing the world's oceans for a cooler Earth,
http://www.exn.ca/html/templates/htmlpage.cfm?ID=19990119-52&Parent=Science

Chissolm, Sallie W. – The iron hypothesis: Basic research meets environmental policy,
http://earth.agu.org/revgeophys/chisho00/chisho00.html

Coale, Kenneth - Report On: Southern Ocean IRONEX III Workshop,
http://color.mlml.calstate.edu/www/news/chismso.htm

Matthews, Ben - Climate Engineering : A critical review of proposals, their scientific and political context, and possible impacts,
http://www.uea.ac.uk/~e256/cleng/cleng.html#top

National Science Foundation - Iron "Fertilization" Causes Plankton Bloom; Scientists Link Iron To Climate Change,
http://www.geo.nsf.gov/adgeo/press/pr9655.htm

Stoll, Michel – CARUSO: CARbondioxide Uptake by the Southern Ocean,
http://kellia.nioz.nl/projects/caruso/

Terry, Kathee - Ocean Color Viewed from Space,
http://athena.wednet.edu/curric/oceans/ocolor/index.html

Trull, Tom Et. al. - Uptake of Atmospheric Carbon Dioxide by the Southern Ocean,
http://www.antcrc.utas.edu.au/antcrc/biogeochem/trull1.html