In case you are now exhausted by the topic of salt in the oceans, I want to reassure you: this is the light at the end of the tunnel.Thus far, I have tried to examine closely and honestly the methodology of Austin and Humphreys (1990), which I described as unscientific and oversimplified. By no means is this a personal attack, as I have documented precisely how Austin and Humphreys have ignored or miscited key data and thus employed unjustifiably simple models to convince readers that the oceans must be younger than 62 million years. They tout confidence not shared by the very authors they cite. Furthermore, they have been resilient in the face of criticism, refusing to update their model despite that more research is available every year, which could drastically improve it.In the last post, I focused on the various mechanisms by which sodium is added to the world’s oceans. By reading through all sources cited by Austin and Humphreys, as well as newer studies from the past 24 years, I found numerous flaws in the ‘sodium inputs’ reported by Austin and Humphreys and utilized in their model. Most of their figures far overestimated the amount of sodium carried to the oceans, and some of the proposed mechanisms add no sodium whatsoever on geological timescales. These authors are thus guilty of some basic errors in accounting, as well as some basic misunderstandings of geochemistry, for which they ought to be held responsible. But that is the nature of science: we open our research to criticism, by which it might be refined. If we refuse to accept that criticism, science cannot advance.
In this final post, I will briefly address the ‘sodium outputs’ reported by Austin and Humphreys (1990), followed by a ‘balanced checkbook’ of the global sodium cycle that shows why the oceans are not missing salt.
|Table 2 from Austin and Humphreys (1990), summarizing
model outputs of Na from the ocean. Units are in 1010 kg/yr.
1. Sea spray
One of the most active and constant processes by which salt is removed from the oceans is felt by anyone that spends much time at (or lives near) the beach. Rust and corrosion are constant worries for any machinery exposed to the sea breeze, which is full of salty droplets of water. Austin and Humphreys (1990, p. 5) describe sea spray rather well:
“Waves of the sea, especially breaking waves along the shore, produce air bubbles in the water. Collapse of these bubbles shoots into the air droplets of seawater which evaporate to form microscopic crystals of halite. Crystals of halite are carried with other aerosols by the winds from the ocean to the continents.”
While sea spray does remove massive quantities of salt from the oceans (Austin and Humpreys estimate 60 million tons/year of sodium, Table 2), the vast majority of this salt returns swiftly to the oceans via rivers and groundwater. You may recall from last post, I likened the process to withdrawing $20 and immediately re-depositing the money into your account. Since I determined the long-term sodium input from sea spray to be 0 tons/year, we must also remove sea spray as a sodium output.
By comparing Table 2 to Table 1 from Austin and Humphreys, we find that sodium lost via sea spray is greater than sodium gained by ~5 million tons/year. I cannot say whether this imbalance was intentional, but it may reflect a real, albeit minor, long-term loss of sodium to the continents. For example, some sea spray particles will end up falling as rain over the Great Basin of the United States, but no rivers drain from the Great Basin to the oceans. In other words, these bits of salt will eventually get buried in sediments or groundwater storage on the continent, providing a long-term sodium sink.
From a ‘deep-time’ perspective of geology, those continental reservoirs of sediment and groundwater may eventually be uplifted and eroded into the oceans. Therefore, it becomes impractical to calculate precisely how much sodium is lost, long-term, via sea spray; we simply know that the amount should be greater than zero.
For the purpose of this discussion, I will follow Holland (2005) and consider the long-term sodium output via sea spray to be 0 tons/year. But we might allow the sodium loss to continents to be as high as the minimum imbalance from Austin and Humphreys (1990), which is 5 million tons/year.
2. Ion exchange
Cation exchange is a blessing to those with ‘hard water’, as water softeners work by exchanging calcium and magnesium for ‘softer’ ions like sodium. In the oceans, the process works in reverse: clay minerals tend to absorb sodium while releasing calcium and magnesium back into the oceans. Since clay minerals are abundant as suspended particles in river water, the rivers deliver millions of ‘sodium-absorbent’ sponges every year.
Austin and Humphreys cite a handful of studies that attempt to estimate the total uptake of sodium via cation exchange. These estimates have not change substantially in recent years, and Holland (2005) uses the same figure (35 million tons/year) in his table. Of course, the total amount depends strongly on the amount and composition of sediments delivered to the oceans, which means that it will vary on geological timescales with riverine inputs of sodium. Therefore, we can take the flux used by Austin and Humphreys as a reasonable, if not a high-end, estimate of sodium lost via cation exchange: 35 million tons/year.
3. Burial of pore water
Marine sediments dominated by clays in particular are extremely porous, meaning that abundant seawater is present between the particles. In short, the seawater gets buried within the sediments, along with the salt it contains. Austin and Humphreys cite an earlier, rather crude estimate of sodium loss via pore-water burial of 22 million tons/year.
We should note that pore-water burial is a complex process, accompanied by numerous chemical reactions (e.g. Scholz et al., 2013). Therefore, it is difficult to estimate precisely the total flux of any element, let alone sodium. In addition, the rate at which marine sediments are buried will vary on geological timescales, depending on the rate and character of global tectonics. During the formation of major mountain belts (like the Andes, Rockies, and Sierra Nevadas), we should expect greater rates of sediment accumulation and pore-water burial, in particular because such mountain ranges are accompanied by deep-water ocean trenches, in which miles of sediment accumulate relatively ‘rapidly’.
4. Halite deposition
Austin and Humphreys’ assessment of halite (NaCl) deposition is rather misleading. They note correctly that modern marine sediments are “nearly devoid of halite”, but do not address completely why this would be characteristic of Earth history. Halite deposition is limited by the fact that halite (NaCl, or ‘table salt’) is extremely soluble in water. For seawater to precipitate NaCal typically requires that a body of seawater become isolated from the oceans, after which an evaporative basin forms under intensely arid conditions. One example of this phenomenon in relatively recent geological history is the Mediterranean Sea, under which thick deposits of salt are buried under younger sedimentary layers. The Natural Historian blog on this topic provides an excellent graphic description of the process and these Mediterranean deposits.
In their discussion, Austin and Humphreys (1990) do acknowledge the existence of such halite deposits in the geologic column, but do not consider it to be a significant sodium sink. To establish this, they divide the global inventory of Phanerozoic halite deposits (4.4×1018 kg of sodium) by the length of the Phanerozoic (they use 600 million years) to produce a ‘time-averaged estimate’ of sodium loss through halite deposition: 7.3 million tons/year of sodium. This number is much smaller than other fluxes of sodium to/from the oceans, so they proceed with confidence (p. 8):
“…it is extremely unlikely that the “time averaged” halite output contains a significant error. No major quantity of halite in the earth’s crust could have escaped our detection.”
Austin and Humphreys derive their estimate of global halite deposits from an earlier study by Holland (1984). Now, what might have changed since 1984? For one, the ability of salt deposits to prime crude oil for harvest has made them a valuable target for petroleum exploration in recent decades. Hence we know far more now about the extent of halite deposits than we did 30 years ago.
As it turns out, the global inventory of halite deposits (~32×1018 kg; Hay et al. 2006) is ~3 times larger than the estimate used by Austin and Humphreys. Based on fluid inclusion analysis and mass balance calculations, Hay et al. (2006) further estimate that about 50% of halite has eroded back into the oceans over the course of the Phanerozoic (an assumption shared by Austin and Humphreys). According to these data, the time-averaged flux of sodium from the oceans via halite deposition is ~35–41 million tons/year. This figure is close to the maximum flux via halite deposition in Table 2.
|Figure 5 from Hay et al. (2006); the distribution of halite deposition by
geological age over the course of the Phanerozoic. Large outcrops of
Cretaceous (K) aged salt deposits are known from Texas, Mexico,
Portugal, and Spain.
In the last post, I suggested that sodium inputs/outputs via chloride solution and halite deposition should not be included in a long-term model of the sodium cycle, because eventually, these halite deposits will be eroded back into the oceans. Technically, we could remove both quantities from the final table. Having included them, however, we can confirm that the estimated fluxes I’ve provided are consistent with observed data. If, on average, 38 million tons/year of sodium are removed from the oceans via halite deposition, and 17.6 million tons/year of sodium are added to the oceans via chloride solution in rivers, then we can expect that after 600 million years, 12.2×1018 kg of sodium should now be locked up in Phanerozoic halite deposits. Since sodium is 1/3 the mass of halite (NaCl), that makes 36.6×1018 kg of halite. This figure is only slightly more than 32×1018 kg, the documented global inventory of Phanerozoic halite deposits from Hay et al. (2006). Within uncertainty, therefore, my refinement of Austin and Humphreys’ model is accurate for the past 600 million years.
5. Alteration of basalt
Sodium removal via low-temperature alteration of basalt on the seafloor constitutes a relatively minor sink. This rate is dependent on that of seafloor spreading, and so it will vary over geological history, but the total flux is too small to impact significantly the final calculation. I don’t see anything problematic with the figure, so I will keep Austin and Humphreys’ cited flux of 4.4–6.2 million tons/year of sodium.
6. Albite formation
I discussed at length in the last post why albite formation is a significant sink of sodium from the oceans and concluded that 25.3 million tons/year of sodium are removed via this process. This is one of the more significant errors in the model by Austin and Humphreys (1990), who mistakenly supposed that sodium was added to seawater through off-axial vents near mid-ocean ridges.
7. Zeolite formation
This final sodium output is likewise so small, that it will scarcely impact the final calculation. Again, I see nothing problematic with the figure cited by Austin and Snelling, so I will leave it intact.
The Global Sodium Cycle in Perspective
After examining all of the supposed inputs and outputs of sodium to and from the world’s oceans, we can evaluate the argument by Austin and Snelling (1990) through an updated table:
Immediately evident from this revised table is the fact that Austin and Humphreys (1990) significantly inflated and overestimated sodium inputs to the oceans. They accomplished this goal by the selective sampling of literature (some of which was already outdated by the time of their publication) and the use of high-end estimates without reporting uncertainties. In addition, they assumed (sometimes blindly) that these fluxes should stay the same over geological time. In fact, none of them should remain constant over tens of millions of years, given the dynamic complexities of our Earth systems.
The flux of sodium to and from the oceans via these various processes is not extremely well understood, even in the modern day. The processes are complex and must be estimated from limited data. Unfortunately, Austin and Humphreys could not afford to be honest about the nature of geology when it comes to documenting global geochemical cycles. In any case, we may finally put to rest the argument that the ocean’s salt content limits the theoretical age to only 62 million years. Given that sodium inputs and outputs are essentially in balance, this upper limit crumbles entirely and is rendered scientifically meaningless.