SECTION 1. INTRODUCTION

I am writing this article in first person, because it flows smoother. This is a long article and it has many (at last count 43) figures in it. The figures are necessary to properly illustrate the principles elucidated, but even I admit that some could be eliminated. I could not decide which ones, so they stay. This study came about when I was preparing to write the article that I have promised my website readers. That promised article was to look at the effects of carbon dioxide with no consideration as to its green house effect. To properly do that, I took a look at where we are currently with our carbon dioxide problem. One of the first figures that I prepared was similar to that shown in Figure 1.

At first glance there is nothing unusual about Figure 1 until you consider several things. Those things are (1) Why is the carbon dioxide concentration higher in the arctic, (Point Barrow, Alaska) than it is for the other three locations? And (2) Why does the concentration go down to a lower level during the summer months in the Arctic than it does for any of the other three locations? , and (3) Why does the lowest concentration occur at different times for the four stations shown? With these questions in mind, I began to look at the available carbon dioxide data and started to sort out causes and effects.

This study has no new data in it. The data that were used were all provided by tw0 organizations (references 1 & 2). This is just a data manipulation study. The “catchy” title is to generate enough interest so that you will read the results and my interpretations. The conclusions drawn from the study are mine alone. Your conclusions might be somewhat different, but it is hard to argue with the overall trends observed. The conclusions, if proven to be correct, will have significant influence on the way in which we approach our carbon dioxide problems, and most importantly, our understanding of the way in which our environment is responding to our carbon dioxide problems.

SECTION 2. CARBON DIOXIDE DATA AND COLLECTION STATIONS

First, I would like to thank all of the individuals that have been involved in the manning of the stations, collections of samples, analyses of the samples and then evaluation of the data. From what I can discern, we are talking millions of man-hours and probably billions of dollars to obtain these data. I can not imagine living and working in some of these locations, particularly the polar stations.

Second, I would like to stress that I have not altered any of the data. When I started the study I made a decision to accept the data as 100% correct. I have also not commented on the reliability of the data, because I have absolutely no way to evaluate the same. The only thing that I did do initially, was to add data where the data were ninety nined (where the data were considered not reliable, the data is replaced with 99 in the appropriate space). This was primarily so that my statistical analyses would come out correct. Later (less than half way through), I found a way around this problem and stopped entering “fill-in” data. None of my analysis has these fill-in data, and I have gone back and hopefully removed them from all of the data plots that I could find.

I gleaned digital data from all of the sources and then hand entered it into my analysis programs, so there may be some incorrect data entry. I plotted all of the data in various ways, so if there appeared to be an incorrect entry, I checked it and corrected if necessary. I work entirely by myself so I have no one to look over my shoulder. If you find some obvious errors that make a difference on the conclusions, please notify me via my web site e-mail address.

I used the data from 34 collection stations all across the globe. There was no particular pattern initially, but as the study progressed, I included stations that would fill in sections of the globe. The 34 stations are shown in Table 1, along with lots of other data that I will discuss in detail later as it fits into the flow of the presentation. There is a 35^th station that was used only to look at the Southern Ocean but was not used in the other data analyses.

The first 5 columns are of interest now. Column 1 is the designated station number. I numbered them according to their latitude, from the north pole to the south pole. The second column is a short title for the station. Column 3 is the latitude expressed in degrees, and the first digit of the minutes(a little unorthodox, but it works). The negative on the latitude designates that it is in the southern hemisphere. Column 4 shows the longitude in the same fashion as the latitude. Here the negative designates the eastern hemisphere. The fifth column shows the height above seal level where the sample was collected. I originally thought this to be important, but later ignored it. The blank for the height above sea level for Baja, Mexico(station 19) is because I never found that data. The last five columns are data that were generated during the study and will be explained later. Try your best not to spend too much time looking at those last five columns now----they will come clear later.

It is helpful when analyzing the data and when you are looking at the graphs, to be able to visualize where on the earth the stations are located. I have therefore included Figures 2 and 3.

Figure 2 is a reproduction of the NOAA CMDL map that is supplied with their data. I have used this map, even though it is not very clear, because it has the names of most of the stations on it. I have added a circle around the location of the station and added the corresponding station number from Table 1. Because the map is very congested in the USA area, I have copied a Rand McNally map that I have purchased and then added the appropriate station circles and numbers on this map (Figure 3).

SECTION 3 GENERAL TRENDS

You have to begin some place, so I decided to plot the annual average CO2 (I will use CO2 as a short hand for carbon dioxide) versus time for the Mauna Loa, Hawaii station. This is generally considered the holy grail of stations and is the most quoted. That plot in Figure 4 illustrates something that is not often discussed.

There are significant breaks in the curve. Most often these breaks are glossed over and presented as just part of the curve. Such will not be the case in this study.

The Mauna Loa data appears to have three or four distinct portions of the curve. The first break appears to occur about 1965, the second about 1976, and the third about 1994. Mauna Loa is the only station that I could find enough data to find the three break points, but all of the other stations had similar breaks. Most of the stations did not have enough data to show all three breaks, but almost all of them had enough data to show the 1994 break. The only station that did not was Mould Bay, Canada, station 3. The two data points after 1994 were not reliable enough to define an entire line.

These data are also shown for Niwot Ridge, Colorado in Figure 5 and the South Pole in Figure 6.

You can argue with the breaks at 1965 and 1976, but the break at 1994 is very dramatic. I have shown only three stations (the three for which I have data including 2009) and the linearity of the plot after 1994 is very good through 2009. I do not have 2010 data yet.

The slope of those lines after 1994 is a measure of the rate of build up of carbon dioxide after 1994. The units of that slope is ppmv/year. The data for all of the stations were plotted and those slopes were calculated. The results of those calculations are shown in column 6 of Table 1. The plots do have varying slopes, as illustrated in Figure 7 which is a typical plot for three of the stations.

I plotted the data in column 6 of Table 1 in lots of different ways and finally decided on the plot in Figure 8.

This is one of the most unimpressive plots in this study. It shows that most of the stations have an accumulation rate of between 1.70 and 1.95 ppmv/year. However, 4 of the stations have rates of 2.00 ppmv/year or greater. They are Park Falls, Wisconsin(#8)--the highest--, Wendover, Utah(#13), Terceira Island, Azores(#14), and Sede Boker, Israel(#18).

It turned out that these stations were highest or lowest in most of the data analysis, and at the beginning of the study I spent a lot of time trying to discern why. I finally gave up because I could not prove or disprove any of my hypotheses, and that was good, because the reason became somewhat obvious later. This was the first time where “trusting the validity of the data” paid off.

From an overall viewpoint, the data in Figure 8 represents the balance between the sources and the sinks for carbon dioxide. With that in mind, it is obvious that something in the neighborhood of the 40^th N latitude is causing the sources to be higher or the sinks to be less than for almost anywhere else in the world. We cannot tell from this data, and must use Figure 8 as a warning to keep our eyes open for explanations later in the study. Not everything in this study showed crystal clear delineation.

Meanwhile, referring back to Figures 4,5, and 6, it is obvious that something occurred on or about 1994 that caused the rate of production of CO2 to be increased or the rate of assimilation by the sinks to be diminished. Keep in mind that there is no way to tell which of these processes changed rates, or if they both did, we only know that something changed. We can further more see that this change occurred on a world wide basis and that nothing has caused it to change since that time. The folks who look at this data all the time will disagree with these statements. They are publicizing changes all the time. BUT, for the trends that are obvious from the data, and considering the precision of the annual averages, it is impossible to see significant changes in these rates after 1994.

With that in mind, all of the data plots and trends that were considered in this study were viewed in light of pre 1994 and post 1994 whenever the data would allow. This caused the number of data plots to be larger than I would have liked. I would like to say that I found conclusive proof of what caused the change in 1994, but I only found inferences, which are discussed in SECTION 18.

SECTION 4. SEASONAL DROP OF CO2

Referring back to Figure 1, it is obvious that there is a seasonal drop in the concentration of CO2 for all of the 4 stations shown. In fact, there is a drop for every station. What this seasonal drop means is that during that time of the year the rate of production of CO2 by the sources is less than the rate of assimilation of the CO2 by the sinks. I thought that this drop might be an interesting feature to investigate so I measured that drop for all 34 of the stations. In order to standardize in some way, I decided that the drop should be measured at the time that that station had an average annual CO2 concentration of 370 ppmv. Where did that come from?

In the next section, I present data for the variation of the seasonal drop with respect to the average annual CO2 concentration. Consequently, it was easy to pick a value off the curve and the value of 370 ppmv meant that I could use the data from all of the stations. The results of those values are presented in column 7 of Table 1 and they are presented graphically in Figure 9.

I chose to plot these data points verses latitude because no matter how I plotted them, this relationship stood out.

WOW!!! For the first time we are starting to get somewhere with characterizing CO2 and its relationship to our world. I decided not to fit any curves to the data, but the trend is obvious. For the southern hemisphere, nothing is going on. For the northern hemisphere between 0 and 50 degrees north there appears to be a somewhat linear rise in the CO2 drop as you proceed north. From 50 degrees north, this trend levels off for all of the data stations north of 50 degrees.

What does this mean? To me, at first thought, it appeared that the CO2 is being removed primarily north of 50 degrees north and from the equator to 50 degrees north, there is a simple dilution, with nothing going on south of the equator. This comes from looking at lots of plots of data from unit processes in conducting pilot plant studies in the water and waste water field. I was all set to model this concept, and I even had almost all of the calculations done when it dawned on me that there is another way to evaluate this process. This other way turned out to yield even more discriminating characterization. This is explained in SECTION 6, but by all means do not skip the next section.

Before we go to the next section, let me say that I did a lot of looking at the variation with respect to the longitude, and I present some of it in section 8, but I never had much success finding trends. It may just be the way I selected stations, but I finally just chalked it up to the possibility that the northern and southern hemispheres are very well mixed in the east-west direction. This might be worth further study later.

SECTION 5. DETAILED LOOK AT ANNUAL VARIATION FOR 3 STATIONS

The shape of the annual CO2 versus time curve can make us more aware of how things work, but the shape can also help us to visualize processes. I first plotted the Mauna Loa data for every day of the year since May 1, 2009, but only for 2009. That plot is in Figure 10.

I did not use the data since the first of the year, because I was only interested in the shape of the annual decrease. By the way, this exclusion of the late winter and spring time data caused me to miss an important trend, but I got lucky and picked it up later on another plot. The CO2 concentration decreases in a linear fashion from day 150 to day 279, and then increases in a linear fashion from day 279 to day 365 where I stopped plotting. Notice the missing data between day 345 and 360. This linear decrease and increase is once again indicative of a dilution model with no interaction.

Plotting the data for the South Pole for the same time period, Figure 11, also showed the linear decrease and linear increase, but because of the significant decrease in scale on the ordinate, it is more difficult to recognize.

Keep in mind that you really need to cock your head almost 45 degrees to the left to get a correct appreciation for increasing and decreasing. I really should have plotted two years, but I can see what I want from the one year.

Now, contrast the shape of the curves of Mauna Loa and South Pole data to that of the Point Borrow data in Figure 12.

There are some similarities, but the bold linearity of the fall and rise are not as apparent, although the rise does have a lot of linearity. I spent a lot of time looking at the small peaks that are evident on this curve that were not evident on the other two curves.

The Point Barrow data collection station protrudes about 3 miles into the Arctic Ocean (I am calling it the whole Arctic Ocean rather than differentiating between the various seas). Most of the descriptions of the area around Point Barrow state that the wind blows almost constantly out of the east or east-northeast, but that is not what I found when I looked at the environmental information. I looked at the wind strength, wind direction, the barometric pressure and the change in barometric pressure. What I found was that none of these parameters married up precisely with the small peaks on the CO2 curve. However, it became obvious that whenever there was a change in the weather as indicated by all 4 of these variables, the CO2 peaks occurred. Not being proficient as a meteorologist, I left it at that.

Because of the location and its relationship to the Arctic Ocean, it became apparent that this station is as close to an ocean station as you can get without being on a ship. That turned out to be important in the later analyses. In Figure 13, I have plotted the data from all three stations as a comparison.

The time difference between the least concentration of Point Barrow and Mauna Loa is obvious---about 60 days. The difference between these two stations and the South Pole station is not obvious from this plot, but in general, the South Pole station has its annual maximum in November or December and its annual minimum in February or March. This is more obvious from Figure 1 where the time lag from Point Barrow to the South Pole seems to be about 7 months.

Keep in mind, that through out the entire study I was trying to see if I could determine if, when, and how much of the carbon dioxide from the northern hemisphere was transferred to the southern hemisphere. Some has to be, or at least that is my hypothesis.

SECTION 6. POINT BARROW CO2 DATA AND ICE MELT

You cannot look at Figure 12 very long before it seems to take the shape of another phenomena---the ice melt in the Arctic Ocean. I obtained ice coverage data for the year 2009 from the graph,(reference 3), which is plotted in Figure 14.

The presumption here is that when the water becomes clear of ice, it is available for CO2 absorption. The correct way to perform the analysis is to determine the open water area and then plot the inverse of this data as a function of time. An easier way is to just plot the ice coverage versus time and see how it compares to the CO2 curve. I did this, but to get them close to the same range, I multiplied the square kilometers of ice, ignoring the 6 zeros after it by 2.27 and then added 364 to this value. I did this for all time frames, and this brought the two curves into the same range on the graph and behold Figure 15.

Things are really starting to get exciting now!! However, as much as we might want it to, this does not prove that the drop in CO2 at the Point Barrow station is solely from absorption in the Arctic Ocean. I have another way to do that in the next section. First, I want to have some discussion about the two curves in Figure 15.

The first thing to note is that the ice data is not really ice area. It is labeled sea ice extent. Sea ice extent is determined be looking at satellite data, sectioning the area in question into small squares and then determining the amount of ice in each square. It a square has 15% or more ice in it, the whole square is included in the count. This means that a lot of the squares could have very little ice in them and still be counted as totally iced. The actual ice area will always be less than the sea ice extent. I present this for clarity, but also to remind me not to get too particular about the fit of the two curves.

One should really be satisfied with the fit of the two curves in Figure 15, but I am not. I think that there are trends that have been analyzed and discussed by two separate individuals in the recent literature. Wie-Jun Cai, et al reported on a data sampling and analysis run in the Arctic Ocean (reference 4) in which they found that the deep artic waters were saturated with respect to carbon dioxide. Following the publication of that article, a Nature news squib quoted Jean-Eric Tremblay of the University of Quebec, Canada with the following: (reference 5)

          “Tremblay warns that what Cai and his Colleagues have observed in the Canada Basin

           May not be true for inshore waters, where winds bring nutrients up from the deep

           Ocean to the shelf. These shelves occupy 70% of the whole surface area of the Arctic Ocean.

           What happens offshore is not necessarily the main driver of the overall CO2 budget, he says. I

           Wouldn’t say that the intake capacity of the Arctic Ocean in terms of CO2 is at its

           Capacity. There is one missing component in that story and that is what happens on those

           Productive shelves.”

That is a long way to go to get where I am going, but it makes understanding the relationship between the two curves in Figure 15 much easier. Initially, the two curves are very close, but as we reach the 90^th day, they start to show different shapes. Keep in mind that initially the ice melts along the shores, primarily everywhere except adjacent to northeast Canada and northern Greenland. Go to the satellite composites of the melting ice and see how the ice melts(reference ?). Later in the melting season, the ice starts melting in the deeper water. That is probably what we are seeing in the ice coverage curve between days 244 and 305. The ice is melting in the deeper water, which according to Cai, et al is saturated and so its effect on the absorption of CO2 is not as pronounced and the CO2 concentration begins to rise before the ice coverage begins to rise. After that time, the curves come back together.

That was a fun bit of sleuthing, and now on to the really good part, but first, I must point out a characteristic of the CO2 curve in Figure 12 & 15 that I investigate later. Looking at the portion of the curve from January 1 to about April 30, it is obvious that the CO2 concentration is rising in a somewhat linear fashion, but very slowly. This portion of the curve has some real significance because it is the time of the year when almost all of the sinks are shut off. It is so cold that most of the waters are frozen over in the northern hemisphere and most of the trees are not taking in significant carbon dioxide. Consequently, it would be of interest to see how this varies with different locations in the northern hemisphere. This portion of the study is reported on in section 13. I put it there because this was a last minute sub-study and it was the easiest place to insert it.

SECTION 7. USING CHEMICAL RATE FUNCTIONS TO DETERMING WHAT’S HAPPENING

You have already gotten the idea that I believe that the CO2 removal at the Point Barrow station is a measure of the amount of absorption by the Arctic Ocean. The absorption of CO2 into water is a first order reaction(defined by Henry‘s Law)---IF the pH of the water remains constant or nearly constant. It becomes much more complex than first order if the pH varies greatly. IF the upwelling water that is spoken about in the previous section is highly active, then it is feasible that the water is removed from the surface before the pH can be greatly altered. This also assumes that a strong thermo cline does not get a chance to set up, because we would not expect the constant pH characteristic to persist in that case. By the way, it is my belief that the saturation that Cai measured is either the result of a strong thermo cline or a strong halocline---something that may be more likely in the Artic Ocean, particularly with all of the ice melt. Cai also stated that they measured salinities as low as 24 Kg/cubic meter.

Now, with all that blab, lets take a look at what a first order reaction means. By the way (again), the reaction rate for the dissolution of CO2 in water is on the order of tens of seconds---very fast by geological standards. (I do not have a reference for that, but can find it if you need). Ten seconds is very slow by chemical reaction standards. This means that if we can get an accepting water in the area of the atmosphere, we can expect a lot of CO2 absorption.

A first order reaction rate means that the rate of the reaction is dependent on the concentration of the absorbed material. If we could some how change the concentration of the CO2 and measure the absorption, and it appeared first order, we would have more proof that the CO2 is being absorbed in the Arctic Ocean. By this time, you are probably way ahead of me----Nature has already changed the concentration of CO2(some people firmly believe that man is primarily responsible). All we have to do is find some way to measure the amount of CO2 that was absorbed during that change. And, of course, we have it in the CO2 drop when the ice is gone.

The proper way to determine the amount absorbed is to measure the area above the CO2 curve from its maximum to its next maximum. Hooray---that’s a lot of work, so lets do it a lazier way. Lets just look at the CO2 drop from the maximum to the minimum. I could not think of a more scientific term than the CO2 drop, so that is what I will call it in the rest of this study.

I did that for all of the data I had for Point Barrow. I looked at the drop with respect to time first and I got a straight line so that was encouraging. Consequently, I plotted the drop with respect to the average annual CO2 concentration and that is what you see in Figure 16.

There are only two data points out of trend (340 & 387 ppmv). The one at 380 is out of line because of the method that I used for calculating the drop. It broke down in 2008 and I did not want to spend the time to do it another way. I do not have an explanation for the point at 340. Remember, at this point, I have spent hundreds of hours. However, it was all worth it and the real meat of the study is still to come.

The fit of the data to our model is phenomenal and the correlation coefficient is 0.99924, even with the two erratic data points. I have drawn the line on the graph that is related to the statistical data and then I have drawn a line that corresponds to Henry’s Law. By the way (again), Henry’s Law is the first order reaction mentioned previously. The question then comes up as to why they disagree. Keep in mind that this variation in CO2 concentration occurred from 1974 to 2009. During that time (the summertime), the open area on the Arctic Ocean has increased almost every year. The absorption of CO2 exceeds the Henry’s Law prediction and I suspect that it is because the area of open ocean is increasing. I will leave it to someone else to refine this process and take into account the increasing ice melt.

I would like to say that this proves conclusively that the annual drop in CO2 at Point Barrow is entirely due to the absorption in the Arctic Ocean, but there is a problem with making that statement. The measurement of CO2 at Point Barrow is the net of all of the sources and all of the sinks. The sources do not magically turn off when the drop begins, so we cannot say with certainty that all of the drop is caused by absorption. With respect to other sinks, I have been told that there are no trees within 200 miles of Point Barrow, and the summers are remarkably brief and cold. However, it still gets above freezing, so the surrounding soil could be either a sink or a source. In addition, when you start looking at the way in which the atmospheric circulation functions in the Arctic and how it is connected to the rest of the northern hemisphere, the surety of that statement gets even worse. The positive aspect to proving the above statement is that most of the wind comes from the north pole, across the ice and then the open water, so its interaction with any other source or sink is doubtful.

What this means is that we cannot say for sure----but the preponderance of evidence indicates that the primary reason for the CO2 drop is because of absorption in the Arctic Ocean. The preponderance of evidence gets larger when we look at rate evaluation for all the other data stations, which is reported in the next section.

SECTION 8. CONSIDERING RATE REACTIONS FOR THE REST OF THE STATIONS

Now that we have another analytical tool, it is only appropriate that we use it. I performed rate reaction calculations on all of the rest of the stations in this study except for Ragged Point, Barbados (#25).

In general, none of the stations have data that is as well linearized as Point Barrow. Some of the Arctic Ocean stations came close, but all of the rest of the stations were so ill defined that it was necessary to used statistical analysis to correctly determine the fit of the data to a straight line. The data for Mauna Loa and the South Pole are shown in Figures 17 and 18 respectively.

The data might be best described as a bloody nose sneezing on a sheet of paper. But even in this case, and particularly in this analysis, no slope or a negative slope are significant data. No slope means that there is not enhanced absorption or removal with increasing concentration and a negative slope indicates that what ever the phenomena is, its removal capacity is getting worse as the concentration increases.

However, when I took another look at the data while I was preparing the final graphs, there seemed to be some type of trend. Take a pencil and connect the data points in sequence for Figures 16, 17 & 18. I did this and then went through the data plots for all of the other stations. I even added data for the 4 stations where I had more data. What I found was that there appears to be a good correlation with the Pacific Decadal Oscillation (PDO), a somewhat poorer correlation with the Atlantic Multidecadal Oscillation (AMO) and even a weak correlation with the global atmospheric temperature. I spent three days on this, and finally just threw up my hands in frustration. This is a complete study in itself. What I did find that is encouraging is that virtually all of the stations had peaks at the following years: 1981, 1984, 1987, 1994 and 2001. I also found that if I moved the Mauna Loa data ahead one year, it matched up much better with the Tutuila, American Samoa data. This kind of approach could lead to finding what I will term the “base station” , the one from where the trends begin. I hope that someone else will follow up on this because I think that it will add a lot of valuable knowledge to our understanding of our CO2 problem.

The rate reaction data are summarized in Table 1. Column 8 shows the slope of the line, which is the most important piece of information. Column 9 shows the intercept, which in my mind has no value---someone else may come up with a way to use it. Column 10 shows the correlation coefficient. In general, after looking at the plots, I would say that a correlation coefficient of less than 0.98 means that we are looking at a sneeze.

It is interesting to note that there are 10 slopes that are negative. One of them is Shemya Island, Alaska. Originally, I considered this one of the Arctic Ocean stations, but when the slope proved to be negative, I concluded that this station is just a COPY-CAT station. What ever concentrations of CO2 that came to it on the wind are sent on with no diminishment. The negative slope could even mean that the CO2 is coming out of solution. I wanted to make a composite for all other Arctic station besides Point Barrow and see how they compared to Point Barrow. Because of Shemya Island’s status as a copy-cat station, I excluded it from my consideration in the Arctic Composite.

All of the other Arctic stations had rate reaction slopes of about the same value as Point Barrow except for Alert and Iceland. I expected the lower rate for Iceland for a host of reasons, but was surprised by Alert. It is my postulate that the amount of ice melt in the Alert area is so low that it diminishes the absorption of CO2---refer again to the ice melt composites (reference 6). I have no explanation as to why Norway’s slope is so high, but there are only 8 usable data points and if you discard the high and the low, then the slope is very near to zero. This is just not enough data (not enough data points) to spend any more time on it.

So, I proceeded to produce a composite of the other Arctic stations. This composite includes Norway, Alert(I put it back in), Mould Bay and Cold Bay and is shown in Figure 19.

This composite was supposed to show a way to account for all of the Artic, but instead it just showed what happens sometimes when you combine data from several stations. The slope of the composite is lower than the slope for all of the individual stations. The data just happened to fit together wrong, or alternatively, there is not enough data for good evaluation of the individual stations (reference the Norway station as discussed above). The arctic composite shown in Figure 19 more nearly represents what I would expect if Point Barrow is excluded, which it is. There is no question that the Point Barrow data are the most consistent with the Henry’s law trend of any of the other 33 stations in this study.

The final step in this section is to plot the slopes of the various stations versus their latitude. That is shown in Figure 20.

There seems to be three ranges. There is the Arctic Ocean stations, all positive except for Shemya Island. There is the southern hemisphere stations, all with low slopes except for Asscension Island (#28) which is located in the middle of the south Atlantic. And then there are the stations in the area of 25 to 50 degrees north. Three of these stations, Wisconsin, North Carolina and Oregon are very high whereas Niwot Ridge in Colorado is not only low, but it is negative. I would have expected these four to be in the same range. At this point in the study I had no explanation for these anomalies, but later I have an explanation that makes some sense but may disagree with the normally accepted norms for atmospheric circulation. It is safe to say at this point, that the middle north latitude stations are all over the place.

It is also safe to say, based on this analytical technique that all of the stations south of 20 degrees north are sound asleep. With respect to CO2 absorption out of the atmosphere, they are all much less active than are the Arctic Ocean stations. It is almost as if the areas south of 20 degrees north have a lid on them. This conclusion will be re-enforced later with another analytical look.

SECTION 9. PACIFIC OCEAN

The data for the Pacific Ocean are presented in Figures 21 through 25. Figures 22, 24, and 25 are data plotted versus longitude. The graph looks weird because rather than figure out how to get Cape Grim, Tasmania, Australia into the chart with it’s east longitude, I just added it’s 40 degrees that it is west of the 180 degrees west to get 220 degrees. It makes the plotting easier.


 

I was expecting to see dramatic differences in the data, both from a north-south and an east-west viewpoint. I was very disappointed that this did not appear. I was so disappointed that I did a very poor job of interpreting this data and I only include it so that my readers can perhaps find the data useful. The stations included Sand Island (#20), the two Hawaii Islands (#22) (#23), Guan (#24), Christmas Island (#26), Tutuila, American Samoa (#29), Easter Island (#30) and Kermandec Island (#31).

The major trend to notice in the Pacific Ocean is the differences and comparisons of the rate reaction data. Two of the stations had negative slopes which is an indication that those stations are less able to absorb CO2 as the concentration increases. These two stations are in the North Pacific. By the way, you could also consider Shemya Island to be in the North Pacific and it too has a negative slope. One could also say


 

that the negative slope is an indication that these stations are releasing CO2 to the atmosphere, but it is not possible to say that with certainty because of the character of the rate reaction data. Some people will find the negative slopes in the north Pacific Ocean to be very interesting because of the interplay of the hot Alaska counter current to the generally hot North Pacific gyre. Once again, that is a study by itself.

The other stations in the Pacific Ocean have rate reaction data that is between 1/5^th and 1/15^th that for Point Barrow. This means that there is very little reduction of CO2 in the Pacific Ocean. The Pacific Ocean seems to have what I call a lid on it.

SECTION 10. A LINE OF STATIONS FROM ALERT TO EASTER ISLAND

Through out the study I was looking for indications of CO2 absorption at the various stations and CO2 transport across the Intertropical-Convergence-Zone (ITCZ) from the northern hemisphere to the southern hemisphere. This section is a report of following the stations from Alert to Easter Island. I selected the year 1996 to study simply because I could get data for all of the stations for that year. Early on, I found that four of the Arctic stations had almost identical shapes for their yearly CO2 curves. These were Alert (#1), Mould Bay (#3), Point Barrow (#4) and Cold Bay (#6). Consequently, I averaged these four data points for each month of 1996 and called this the Arctic Composite. That composite curve along with Shemya Island (#7), Sand Island (#20) and Cape Kumukahi (#22) are shown in Figure 26.

Shemya Island is still acting a lot like the Arctic Composite even though it has previously been branded a copy cat because of it’s negative slope on the CO2 drop vs. average CO2. This illustrates that the various data stations can exhibit various characteristics. The general trend for the data in Figure 26, and for all of the stations is that as you move south in the northern hemisphere, the following things happen:

          1. The maximum CO2 concentration in the first three months of the year decreases.

          2. The peak that occurs about the 5^th month gets pushed to a later time.

          3. The bottom of the drop diminishes.

          4. The bottom of the drop gets pushed to a later time.

Moving on to Figure 27, I have once again shown Sand Island and Cape Kumukahi but have added Mauna Loa, Christmas Island and Easter Island.

Keep in mind that Easter Island is 27 degrees into the southern hemisphere. You can compare the two stations at Hawaii. Cape Kumukahi is collecting samples at 3 meters above sea level whereas Mauna Loa, at almost the same location is collecting samples at 3397 meters above sea level. Mauna Loa’s peak is a little lower, its valley is a little higher, and the drop occurs at a slightly later time (only a matter of days). This is to be expected as the Cape Kumukahi station is picking up CO2 from the islanders and the vegetation is absorbing some of the CO2. This type of relationship has been reported in several places in the literature (I will provide references later if requested).

Before moving on to the Christmas Island data at 2 degrees north latitude, it is instructive to observe where the maximum and minimum CO2 concentrations occur in the southern hemisphere. Referring back to Figure 1. It can be seen that the peak concentration for the South Pole occurs in November or December whereas the minimum CO2 concentration for that year occurs the next year in February or March. Obviously, this is because of the opposite summer and winter seasons from the northern hemisphere.

Now, back to Christmas Island. It is obvious that there are two peaks and minimums for the Christmas Island data. The first peak at the 3^rd month is about half way between the northern hemisphere peak and the southern hemisphere peak. The first drop is minimum at about 1 month earlier than the minimum for Easter Island . Christmas Island then displays the expected peak between the 6^th and 7^th month with the minimum between the 9^th and 10^th month. This is rather a rather powerful argument that the Christmas Island CO2 concentrations are being strongly influenced by the southern hemisphere. This is a more visual demonstration of the southern hemisphere’s influence on the northern hemisphere than I would have expected. However, keep in mind that this is only a 1 ppmv drop, or thereabouts, for about 2 months. Also keep in mind that the Christmas Island sample collection site is only 3 meters above the sea level. Ordinarily, I would expect this low sampling location to mask the transport of CO2 higher in the troposphere. This means that the change in concentration higher in the troposphere may be much larger, or alternatively, that the data reflect another phenomena at the 3 meter height, such as local absorption. Referring to column 8 in Table 1, we see that the CO2 uptake rate is positive and three times the rate at Easter Island, but only about 1/5^th as large as Point Barrow. I cannot say for sure that the dip in the Christmas Island data is definitely a function of mass transport across the ICZ, but it appears to be. Even if it is transport, it is very small.

Referring back to Figure 27, it is not obvious that the Easter Island data shows any sign of CO2 transport into the southern hemisphere. Consequently, I performed another study as reported in the next section.

SECTION 11. DETAILED STUDY OF EASTER ISLAND DATA

I plotted out the monthly average CO2 concentration for Easter Island for the years 1994 through 2002. I thought I could see some evidence of a northern hemisphere peak---remember that Easter Island is 27 degrees into the southern hemisphere. I decided to compute a composite for that 9 years of data. Sometimes a composite brings out traits you are looking for, and sometimes it hides them. When I did the composite, shown in Figure 28, I looked carefully at the data to see if any missing data would bias the result.

I did not see any. Consequently, I have more than the average faith that the small peak that I have noted on the Figure is probably real. At best, you could say is that the small peak is indicative of mass transport of CO2 across the ITCZ, just as the same is true for the Christmas Island data. At worst, you could say that the procedure is faulty and the data should be disregarded. Even if you buy into the data manipulation process, it is obvious that the amount of CO2 transported is very small. I thought it was worthwhile or I would have not presented it. I will present another technique in a later section that tends to support the transfer yet shows that it is not significant.

Why am I spending so much time and effort on this question? It is because I am developing a sense (I have not been able to quantify it) that the major transport of CO2 from the northern hemisphere is not via the atmosphere, but is rather by the oceans. More on this later.

SECTION 12. DETAILED STUDY OF CHRISTMAS ISLAND DATA

I spent a lot of time on the Christmas Island data because it appears that something is happening in that area. This would be expected because this area is integrally involved in the development of both the el ninos and the la ninos. I have spent a lot of time looking at the sea surface temperatures in that area and it appears to me that the el nino is lit off from two separate heat sources just off the coast of Costa Rica. That too could be expected because there are several tectonic plates that meet in that area. As I looked at the CO2 data in that area, as represented by Christmas Island, I found something that I had been looking for elsewhere but had not found.

I plotted out the CO2 data for Christmas Island from 1973 to 2000 in six year groups. The reason for the six years is that my plotting program only holds six separate graphs. I noticed that the 1973 to 1978 data looked very similar to the 1984 to 1989 data and both of these had some stricking variation with the 1994 to 2000 data. Consequently, I averaged the three six year data for every month. I then added 15 ppmv to all of the monthly data for 1973 to 1978 to get it in the same range as the middle years data. In a similar fashion I average the other group and then subtracted 15.5 ppmv from this later group. These results are shown in Figure 29.

If you do not approve of my averaging technique, then skip this section. The thing of real interest in Figure 29 is the close agreement between the two earlier averages---they fall almost entirely on top of each other. The other thing of interest is the dramatic disagreement with the later averages. There was a fair amount of missing data in the 1994 to 2000 data, and so I inspected where the data were missing. At months 1, 11, and 12 the data were not only missing, but they were missing from the upper or lower values of CO2 which unduly biases the data. For this reason I have indicated the number of data points at each month. I consider the first and the last two data averages to be heavily biased. I consider the other data points to be representative because the missing data point or points were in the middle. In addition, the data point at month 5 is on a line with the data points for the month before and after.

The data in Figure 29 are related to my search for a cause for the change in the slope of the CO2 versus time plots at virtually all of the data points. Something is quite different in the data from the pre-1994 and the post-1994 areas. For the 1994 to 2000 data, the rise in the 2^nd through the 4 month is faster, there is a dip in the 4^th through 6 month that was only hinted at in the 1973 to 1978 data, but the overall slope of the decline from the 4^th through the 9^th month is about the same. The accelerated rise in the 2^nd to 4^th month is to be expected, but the dip in the 4^th through 6^th month is very pronounced and may have some significance. Once again, there is nothing absolute, only an accumulation of a preponderance of information.

SECTION 13. RISE OF CO2 DURING THE WINTER MONTHS IN THE NORTHERN HEMISPHERE

As mentioned at the end of section 6, the rise of CO2 during the winter time in the northern hemisphere has special interest because it is a time when the sinks for CO2 are not very active---perhaps entirely inactive. Consequently, investigating that range is of interest.

The procedure for this portion of the study was to observe the average CO2 concentration for the month of January, enter it in a plotting program, then observe the maximum CO2 concentration in March, April or May and then enter that in an row for the same year. These two values were then subtracted to get the CO2 rise for that year. This was done for all of the years for which there was data available for that station.

These resultant data were plotted versus the year for 6 or 8 stations in the northern hemisphere (These plots are not shown----yes I had a very large collection of plots that were not used in the final report). It was observed that for some of the northernmost stations, there was a slight downward trend of the data with time, but for most of the stations, the rise of CO2 appeared not to vary much with the year. For this reason, I decided to just average all of the data for every station.

The results of those analyses are presented in Figure 30.

The first thing to notice is that there is a negative value. This is for Grifton, North Carolina (#16) which station has only 6 years of data. The value came out negative and this doesn’t make much sense, but I left it in. I did not use the Christmas Island data even though it is in the northern hemisphere. We already know that this location is heavily influenced by the southern hemisphere.

The next thing to notice is that the ordinate ranges from about 1 to 4 ppmv. This is the build up of CO2 during the time that there is virtually no removal----WELL, we must qualify that: there is virtually no removal for those locations where the world is frozen. I have presented data for areas where it is sunny and warm all year as well as for stations where it is frozen in the winter time, so it is necessary for us to separate the figure into two zones to get a proper analysis of what the figure means.

I have arbitrarily said that anything north of 37 degrees north qualifies for the frozen world during the time of January through May. This means that the increase in atmospheric CO2 during this frozen time is between 1.0 and 2.5 ppmv. That is not a very large increase and it would be nice to see how that fits in with the generation of CO2 by the sources. I’ll do that later, but first I want to discuss the rest of the curve in Figure 30.

To properly evaluate the significance of the rise of CO2 during this time for the mid and low latitudes for the northern hemisphere, we must refer back to the data in Figures 1, 26, and 27. As we proceed south in the northern hemisphere, the CO2 peak occurs later and later in the year. This means that there is a larger difference between the January data and the maximum point. This means that by the vary nature of the way in which I generated the data for Figure 30, I biased the data. I recognized this before I began, but it still seemed to me that there would be some informative data in the trend.

I decided that I would not use any of the data for latitudes less than 37 degrees north, but I would use the trend to make an inference on the data north of 37 degrees north. I inferred that there would be a rise in the amount of CO2 removed for the data north of 37 degrees and therefore it would be O.K. to use an average for the data to get a representative CO2 rise for that whole area. Consequently, I used an eye ball average and said that the CO2 rise for that area is 1.75 ppmv for that 3 month period. I will use that figure to calculate an annual CO2 increase and then compare that with the measured increases for all the stations.

The reasoning goes as follows:

          -1- We must multiply the 1.75 ppmv by 4 to get an annual rate=7 ppmv/yr.

          -2- The area north of 37 degrees north represents 19.6% of the globe’s area

          -3- Multiply 7ppmv by 0.196 to get 1.37 ppmv for the entire globe per year

This compares to the average annual increase of about 2 ppmv/year.

That is not a very good agreement because the 1.37ppmv is supposed to represent the situation when there is no removal of CO2 from the environment whereas the 2 ppmv represents the real situation in which about 53% of the CO2 that we put in the air stays in the air (reference 7).

Considerations that make the figures not very realistic are: (1) There are lots of sources that lie outside of the 37 degrees north influence, (2) Some of the CO2 is transported out of the area during this winter time, (3) There are no considerations given to removal by the rest of the northern hemisphere during this winter time and (4) There even appears to be a small amount of removal by the Artic Ocean during this first 3 months of the year.

With all of these negative considerations, it is my belief that the two figures are remarkably close. All interactions in the environment turn out to be incredibly complex and when you can make the simplifying assumptions that I made in these calculations and still come out this close, I am satisfied.

I admit that I am not as excited about the results of this phase of the study as most of the others. However, someone else may see uses that I have overlooked. To me the value is to show that the rise of CO2 when the sinks are not as active is in the ball park of expectations.

SECTION 14. ANNUAL AVERAGE CO2 CONNECTIVITY WITH LATITUDE

This is probably the first data analysis technique that I should have used, but my brain does not always function in a logical manner. I stumbled onto the technique when I reasoned that the annual averages may mask some important traits. Consequently, I plotted the maximum CO2 concentrations for all the data stations and connected the data points for the same year. Because of my concern for missing data, I plotted three years at the same time (I think it was 1995, 1996 and 1997) and the result was amazing.

I am presenting the data plots in what I contend is a logical fashion---present the averages first and then look at the extremes. The averages for the 1983 to 1986 data are not very dramatic, Figure 31,---not because they do not show similar things to the plots for the other groups of years, but because there is a lot less data available. Only 17 of the 34 data stations have data in this range.

Now, look at Figures 32 and 33.

These are data for two groups of 4 years each. Some trends start to appear. Something is going on in the range of 65 degrees north to 85 degrees north but in general, the data are somewhat constant. In the range from 65 degrees north to 45 degrees north, the general trend is increasing as the latitude decreases. In Figure 31, the change in concentration in this range is rather haphazard---down some years and up others. However, for the years 1995 through 1998 the trend gets stronger (Figure 32) and in the last set of years (Figure 33) the trend get dramatically stronger.

Continuing with the trend analysis, the data from 25 degrees north to 45 degrees north is absolutely chaotic, BUT it is uniformly chaotic. From 25 degrees north to 15 degrees north there is a decreasing trend in all of the data and then from 15 degrees north to 2 degrees north, there is an increasing trend. Crossing the equator, there is a decreasing trend and then there are some uniform ups and downs as you progress to the south in the southern hemisphere.

As most of you know, the general atmospheric circulation in the northern hemisphere is ruled by the Hadley cell from the equator to about 30 degrees north, the Ferrel cell from about 30 degrees north to about 60 degrees north and then the Polar cell from 30 degrees north to the north pole. To refresh your memory on the atmospheric circulation cells, see reference 8. The limits of each of these cells moves around with prevailing conditions and particularly with changes in seasons. Because of the way in which these cells function, the Polar cell tends to sweep the surface air in a southerly direction, modified by the coriolis effect to slant the winds to the west hence the predominant wind at Point Barrow is from the east or northeast. The Ferrel cell tends to sweep the surface winds to the north, modified by the coriolis effect and the Hadley cell tends to sweep the surface winds to the south, once again modified by the coriolis effect.

Looking at the data in Figures 32 and 33, one could postulate that the Polar cell is reaching down to 45 degrees, the Ferrel cell is spanning from 15 degrees to 45 degrees and the Hadley cell is reaching from the equator to 15 degrees. The only thing wrong with this postulate is that the Hadley cell has been found to be much the strongest of the three and I do not think that anyone would believe that it has gotten so small. However, if you take the Key Biscayne data out, then the limit of the Hadley cell goes back to 30 degrees and it all gets a lot more believable. I said at the beginning that I would not evaluate the reliability of the data, so I left the Key Biscayne data in.

The next problem is that the Polar cell seldom reaches further south than 60 degrees. This one I will argue against. I believe that the Polar cell is reaching much further south than normal and that the data are representing that fact. It is my belief that the latest data from the AIRS satellite (reference 9) support this claim. However, those data are at 10 kilometers above the earth, so one could argue with the use of that data when comparing it to these surface stations. As I have said above, I am not enough of a meteorologist to be qualified to argue the issue. However, if we look at the maximum CO2 trends which occur in winter and then look at the minimum CO2 trends which occur in the summer, then we might see an expected shift of what I consider to be the confluence of the Polar cell and the Ferrel cell. That is what the next two sections are about.

SECTION 15. MAXIMUM CO2 CONNECTIVITY WITH LATITUDE

The data for this section were handled in the same fashion as the previous section except that maximum CO2 concentrations were plotted rather than average concentrations. The data are presented in Figures 34,35, and 36. Once again, the data in Figure 34 are for years in which the data are inadequate to do more than look somewhat at trends.

However, the data in Figures 35 and 36 are just as sharp as in the previous section. Caution---when you are comparing Figures 35and 36 be warned that the vertical scales are not the same. This alters the appearance of the comparative figures.

The data in Figure 35 reinforces the conclusions from the previous section. The major difference is that there is no rise in the CO2 concentration when going from 30 degrees north to the equator. There is a significant fall in the CO2 concentration. Keep in mind that CO2 is not a conservative tracer although, in the winter time it is more of a conservative tracer than in the summer time.

The barrier at 45 degrees north that is causing the rise in CO2 when coming from the north pole is still there and in Figure 36 it has become even more dominant (that is the reason that I had to change the scale). The difference in the CO2 concentration between the Arctic Ocean and the Southern Ocean is about 6.5 ppmv. In the previous section, using annual averages, the difference in the average was about 3 ppmv. I do not know what that means, but it keeps us in perspective.

The difference between what I term the Hadley cell and the Ferrel cell has somewhat disappeared in these figures. It will be interesting to see what the next sections shows. Based on what we see in these figures, it is appropriate to comment on the very high CO2 values, very high CO2 drops and the very high slopes for the CO2 drop versus ave CO2 curves for the stations at Wisconsin, North Carolina, and Oregon. It seems that these stations are at what I have termed the barrier at about 45 degrees north latitude. Because the CO2 appears to be piling up at these locations, this could explain all of these high values for these three stations.

SECTION 16. MINIMUM CO2 CONNECTIVITY WITH LATITUDE.

Before I begin dealing with the results of this section, it is important that I explain a problem and how I dealt with it. It has significant impact on the data in the southern hemisphere. The maximum CO2 concentration for the southern hemisphere occurs in the year in question (November or December as shown above). However, for the minimum CO2 concentration, the minimum occurs in the next year (February or March as shown above). If you use the minimum for the year in question, it biases the data. This became very obvious when I plotted the minimums. There was a sharp drop at about 35 degrees south latitude that I had not seen before and seemed to be out of character with the data. So the problem becomes where do you switch to the following year for the correct minimum CO2 concentration.

This becomes a problem of determining where the influence from the northern hemisphere effects the data for the southern hemisphere. I solved the problem by making an arbitrary decision that I would use the following year’s data for all points south of 35 degrees south. This makes approximately a 1 ppmv change in the data for the southern portion of the southern hemisphere. However, if you are drawing conclusions from the southern data, be aware of my decision. I am drawing very few conclusions from the southern data.

With that said, peruse the data in Figures 37, 38, and 39.




 

Pretty much ignoring Figure 37 as per the other sections, we see that the barrier at 45 degrees north has moved to 40 degrees north.

Everything else is very much the same as far as trends are concerned. I would have expected that barrier at 45 degrees to have moved north rather than south to 40 degrees north, but I do not have a reason for that expectation.

At this point, I have exhausted my brain on ideas about these three sections of data analysis. However, my curiosity is still alive so I wanted to see what the maximums and minimums for the same year would look like on the same graph. Thus the next section.

SECTION 17. MAXIMUM AND MINIMUM CO2 CONNECTIVITY WITH LATITUDE

I have presented data for three individual years in each one of the time zones. I tried presenting all four years for each time zone, but the graphs got too crowded. See Figures 40, 41, and 42.

The presentation is pretty much the same as for the previous three sections. The one exception is in the southern hemisphere and is obviously influenced by my decision as to how to plot the minimum data. The variation between the minimum and the maximum for the southern hemisphere is less than 1 ppmv for all graphs. That is even the case for latitudes north of 40 degrees south where I did not alter the selection of minimums. To me, this means that the difference between the maximum and the minimum is less than 1 ppmv and yet the CO2 concentration in the whole southern hemisphere is increasing in the neighborhood of 2 ppmv per year. Once again, this begs the question of how is the CO2 getting down there if most of it is produced in the northern hemisphere.

Looking at the data in these figures shows that there is a very sharp decline or increase (max. or min.) in the CO2 concentrations as you progress across the ITCZ in the neighborhood of the equator. You notice that I did not say “at the equator”. It appears that the ITCZ is south of the equator by 5 to 10 degrees. Also, the difference between the maximum and minimum concentrations is less than 1 ppmv by the time you get to 15 degrees south and nothing else seems to happen from then on south. This too begs the question of how is the CO2 getting to the southern hemisphere.

SECTION 18. DISCUSSIONS OF BREAKS IN THE CO2 VERSUS TIME CURVES.

I gave a limited discussion of the break in the rate of increase in the CO2 versus time curve in SECTION 3 and promised more in this section. About the only data that I could generate to illustrate the effect of that break was in the detailed analysis of Christmas Island data, SECTION 12. There was demonstrable differences in this data prior to 1994 in relation to after 1994. Nothing more substantial than that.

Because of the time frame and because the rate of CO2 build up has been so constant after 1994, I surmised that this must be a global phenomena. To determine if it was related to the CO2 emissions by man into the environment, I searched and found data on global emissions by year from 1974 to 2006 (reference 10 ). Those data are plotted in Figure 43 and they show a nearly linear trend from 1974 up til 2002. The slope of that line was 0.303 giga tons per year (a giga ton is one billion metric tons) and shows no indication of change on or about 1994.

However, in 2002 China started generating lots of CO2 emissions and the slope increased by a factor of 3 and has been quite linear for the 4 years after 2002. Neither the lack of change in 1994 or the radical change in 2002 shows up on any of the CO2 versus time plots in this study. I only found 4 stations that had data that recent and none of them showed any deviation from the rate established in 1994. This would tend to indicate that the slope of that line is not dependent on emissions, but some other natural phenomena in the world.

I went looking in the literature and found an interesting article McPhaden and Zang (reference 11) It is entitled “Slowdown of the Meridional Overturning Circulation in the Upper Pacific Ocean” . I have read a lot of technical reports, but I must admit that reading this article is more difficult than translating Russian, a skill that I was once somewhat proficient. If I make some errors in determining what they say, some one please e-mail me and set the record straight.

They looked at an area of the Pacific Ocean from 9 degrees south to 9 degrees north and computed data for the transport of several types of water into and out of that area. They calculated these parameters for 4 time periods, 1956-65, 1970-77, 1980-89 and 1990-99. In summary they found that for the time period prior to 1990, the pycnocline convergence in this zone decreased from the mid 20 to about 14 Sverdrups for the time period of 1990-99. They further found that the equatorial upwelling decreased from the mid 40s prior to 1990 to about 35 Sv after 1990. Pycnocline water is cold water with a specific density. Equatorial upwelling is somewhat self explanatory. The term Sverdurp is a term invented to describe extremely large flows that occur in the ocean. One Sverdrup, abbreviated as Sv is equal to one million cubic meters flowing by a given point per second. The flow in the Gulf Stream off the Florida coast is approximately 50 Sv. The total flow of all the fresh water rivers in the world is about 1 Sv.

If the pycnocline water into the area decreased by about 6 Sv and the upwelling in the equatorial area decreased by a similar amount, this means that there is that much less flow of cold water to the surface. Cold water has a very high solubility with respect to CO2 and this means that there is that much less water available to absorb the CO2 out of the air. As I said previously, the slope of the CO2 time curve is a net amount of the sources minus the sinks. If the sink capacity goes down, this will be evidenced by us as an increase in the rate of rise of the CO2 curve.

To put these flow rates in perspective, the entire flow of hot water into the Artic Ocean is between 6 and 8 Sv, about the same as mentioned in the above article. This hot water is sufficient to account for the majority of the ice melt in the Artic Ocean(see my article entitled “Artic Ice” at this web site for the calculations).

The connection between the research noted in the above article and the break in the CO2 curve is only speculative but the time frame is about the same. Perhaps other researchers out there will take more notice of other correlatable phenomena in the future. I could find no other reasonable explanations in the literature. It was one of my goals when I began this study to attempt to explain those breaks, but I was not successful.

SECTION 19. DISCUSSION OF THE “LID” ON ALL OF THE OCEANS EXCEPT THE ARCTIC

In many of the sections above, I refer to a lid on the oceans of the world except for the arctic. Obviously, I have thought about this lid a great deal as I have been performing this study. It is impossible not to, because the Arctic Ocean, which is the smallest of the five major oceans of the world is apparently removing almost all of the CO2 from the atmosphere. Just looking at the connectivity plots makes this plain, but in this study I have found several other measures of this preponderance of removal by the Artic Ocean. So the question keeps popping up as to WHY are the other oceans not helping?

I have spent a lot of time looking at the results of ocean surveys of CO2 for all of the oceans of the world, and there is one trait that is evident above all of the others. That trait is that as you go from the bottom of the ocean to the top, the concentration of CO2 decreases quite slowly (seldom more than 50 micro moles per kilogram). However, when you get near the surface, that decrease is concentration is very sharp and rapid. By the time you get to the surface, the concentration has gone from typically 2100 micro moles per kilogram to less than 2000.

This rapid drop at the surface could typically be blamed on a couple of things. The first is that the ocean is releasing it’s CO2 as you approach the surface. This might be expected where the water is hotter than normal, but generally, the 2000 micro moles per kilogram is much lower than would be expected for today’s atmospheric concentrations of CO2, if the water is in equilibrium with the atmospheric concentration of CO2. And, of course, in this study we see that when conditions are right in the Arctic Ocean, the ocean just sucks the CO2 up.

The second thing that could cause a lower than expected concentration of CO2 at the surface is what happens when a thermo cline sets up. <<<please excuse my splitting the word thermo cline into two words---my word processor does it automatically and I don’t know how to over-ride it> >> This thermo cline isolates the top 100 meters of the ocean from the rest of the ocean and provides a barrier. This barrier takes lots of forms, but the one that is of interest to us is the acidification of that surface water above the barrier. This acidification (drop of pH) comes about from the increasing concentrations of carbon dioxide; yet we already said that the CO2 is lower above the thermo cline than in deeper waters. Consequently, the acidification must come about from the biological activity in these surface waters, and of course, that is well documented. Values of pH in the surface water of as low as 7.6 (reference 12) are common even when the overall pH of the oceans is at or above 8.1.

I have done some simple calculations, using both the average Henry’s Law constant for ocean water and some simplifying assumptions relating to the speciation of the bicarbonate-carbonate ions in sea water. I used an average atmospheric concentration of CO2 of 400 ppmv (we are close to that now) and I calculated the equilibrium concentration of CO2 in sea water at several pH values. The results of those calculations are shown in Table 2 below: