RealClimate logo

The age of Aquarius

Filed under: — gavin @ 12 June 2011

The international Aquarius/SAC-D satellite was successfully launched yesterday (thankfully!). Media coverage was good – except for the almost absolute avoidance of the term ‘salinity’ to describe the concentration of salts in the surface ocean that Aquarius will retrieve – oh well. But what is Aquarius going to see, and why is it important?

The most important factor governing the dynamics of any fluid is the equation of state that relates its composition to its density. That ultimately determines the pressure gradients and hence the velocity and circulation. In the atmosphere, the important constituents are temperature and water vapour, while in the ocean it is temperature and salinity. Making seawater warmer or making it fresher (by adding rain or river water) both decrease the density and make the water more buoyant. Similarly, cooling and evaporation both make seawater more dense. Salt also affects the freezing point of water (it is around -1.8ºC for normal seawater, colder still for seawater under pressure), so salinity can affect sea ice evolution too.

The specific definition of salinity itself has subtly changed over time. Originally it was the mass of dry solids left over after all the pure water was evaporated (measured in g/kg or similar), but that is difficult measurement to make on a routine basis. Then it was defined by measuring the Cl concentration (closely related since NaCl is the dominant salt). More recently, salinity was measured via conductivity (saltier water conducts electricity more readily) and reported in ‘practical salinity units’ (psu), and in 2009 the official definition changed again to a new ‘Absolute salinity‘ (see the TEOS-10 website for details). All of these are roughly coherent, but there are subtleties that have become more important as knowledge has increased (for instance, related to the (small) impact of varying carbonate concentrations), and realisation that very small changes in density can have important effects.

In the surface ocean, there are strong salinity gradients across the major currents – water is much fresher on the west side of the Gulf Stream extension than on the east side for instance. Warm core eddies in the South Atlantic are saltier than surrounding sea. The western pacific warm pool is fresher than water in the East Pacific since it rains more there. One important thing to note is that temperature and salinity anomalies are often balanced – that is cold water is often fresh, and warm water is often relatively salty. Since these two factors have opposite impacts on the density it is difficult to judge whether water is more or less dense (which is key to the ocean circulation) just from the temperature alone.

This is important, for instance, in attempts to predict ocean circulation based on current conditions – if you only use temperature information (because we don’t yet have good salinity data in real time), then you can incorrectly assume density anomalies that are much too large, or even the wrong sign (this was one of the problems in the Keenlyside et al paper from 2008).

Aquarius will hopefully give a much needed boost to attempts to track ocean density for these purposes. But there are a lot of other variations that will likely be seen that will give a lot of insight into important ocean dynamics. What is the role of salinity variations in the development of an El Niño event? Can we validate rainfall and evaporation estimates by looking at the change in ocean salinity? Can we quantify sea ice melt by its impact on salt? I’m pretty confident that Aquarius will reveal a far more dynamic picture of sea salt variations than we currently imagine.

Aquarius retrievals are based on passive microwave technology and rely on the fact that salinity affects the thermal emission properties of the ocean surface. This effects are quite subtle, and the range of variability is relatively small, so it has taken many years for the technology to catch up to the need. This is however a first attempt to do this from space, so challenges will undoubtedly remain. (Correction: The ESA SMOS mission (launched in 2009) was in fact the first salinity measurement from space.)

As an aside, NASA has a salinity quiz to test your knowledge. Rather unfortunately I only got 7/10 right, which (at least in two cases) I put down to ambiguities in the questions…. but I’ll be interested to see if others do better (we can discuss the details in the comments). There are also some minor errors in the education portion of the Aquarius site (e.g. sea ice has an average salinity of about 5 psu, not zero), so if anyone spots anything else, let us know and I’ll try and get it corrected.

65 Responses to “The age of Aquarius”

  1. 1
    Joel Shore says:

    Re the quiz: I got 8/10, admittedly with the help of the figure in your post here. I agree that a couple of the questions were ambiguous, particularly what one considers to be “a very small variation”.

  2. 2
    Magnus W says:

    Well for starters, what is wrong with the metric system?

    A warmer world = more weathering = more salt in the ocean?

  3. 3
    Don Smith says:

    9/10 here but I meta-analyzed the two very ambiguous questions. I don’t consider 5 ppt range in a 35 ppt value small, but I guessed that the questioners considered 5 ppt a small amount.

    The last question is the one I got wrong. Won’t all that melting of glaciers, ice sheets and sea ice decrease salinity? It seems obvious to me,

  4. 4
    Thomas says:

    The quiz is a complete noshow on my old version of firefox, so I’ll have to stay at 0/0.
    Magnus@2. Average salinity (or rather total dissolved salt inventory) would respond to the difference in erosion of evaporite deposits and deposition of new deposits. What Aquarius is more interested in is understanding the ocean atmosphere interaction: evaporation minus precipitation, and its effect on density/ocean circulation. Of course melting if ice sheets and sea ice does mean there is more volume of seawater to contain the inventory.

  5. 5
    Edouard says:

    I did the test with my son Antoine (14) who got used to swim either in the Mediterranean or the Baltic Sea… There is at least a factor two between the salinities which you can even feel by tasting the water. So we got trapped by the question about the relative stability of the global ocean !

    In large portions of the tropical and temperate ocean, the salinity at the surface is higher than near the bottom.

    Over the next century, global sea level will rise by XX cm (or 1YY ?) partly because of freshwater inputs from glaciers and polar ice caps. Shouldn’t the salinity (i.e. global average) inevitably decrease ? (even if gradients will change notably at the surface and even at depth).

  6. 6
    Prokaryotes says:

    “One important thing to note is that temperature and salinity anomalies are often balanced”

    Can we assume that with ice melt this “balance” is likely to become “unbalanced” or rather more pronounced?

  7. 7
    Alex S says:

    Thanks for the great article, but I’m a bit confused about some of your salinity examples – they don’t seem to match the included salinity map. For example, you say the western gulf stream is much fresher than the eastern gulf stream. I don’t see that difference on the map at all – florida looks just as about as saline as the tip of portual. Perhaps you mean the northern extension of the eastern stream? Also, the western and eastern pacific don’t look very different to me. Any clarifications are greatly appreciated!

  8. 8
    Sou says:

    Well I’m not a complete nong – getting 7/10 as well, also fooled by the ‘very small variation’. Didn’t know that the Atlantic was the saltiest ocean (I should have read your article more carefully – I picked Indian for no reason at all except it ‘felt right’!) Nor did I figure out about salty water sinking. (Visiting mostly warm beaches, I figured that evaporation made the surface saltier, but the density makes sense when I think about it. Again, I should have read Gavin’s article properly before taking the quiz.)

    The Aquarius effort sounds fascinating and extremely useful, particularly re the role that ocean currents play on climate. Good on you to all who made it happen.

  9. 9
    Hans Kiesewetter says:

    Re #4 (Thomas): it is not only your old version of Firefox. Also my iPad shows a blank screen. Score: 0 points.

  10. 10
    Magnus W says:

    Thomas 4,

    It was about one of the questions in at the other site.

  11. 11
    Larry Atkinson says:

    I got 9/10 but consider that luck and guessing what they wanted. I got 6 wrong but it is a classic case of asking a simple question about something complex (that goes for most of the quiz). And 1000 pounds! Give me a break. Since the question was just about percentages they could have used kg.

    Regardless – let’s hope the data starts coming in and is of the quality the instrument was designed for.

  12. 12
    CM says:

    8/10, after scanning your post — would have been 9 if I’d paid attention to the map.

    Congratulations on the new bird. It’s an optimistic name; may we yet see the dawning of a climate debate with no more falsehoods and derisions.

  13. 13
    jyyh says:

    6/10 and blaming Indian Ocean for not evaporating more… ambiguities for 2 more… :-) fun quiz though.

  14. 14
    Lorne50 says:

    Got 10out of 10 That was so easy a driller got em right

  15. 15
    tamino says:

    I got 9/10, only missed the last one (I too expected melting glaciers and sea ice to freshen the oceans). That surprised me — since you only got 7/10 I expected to do worse than that. But, I too was helped by your article (which shows the enhanced salinity of the Atlantic.

    [Response: I agree that the last question is probably wrong – the expectation will be that global salinity will decrease (though I could think of circumstances in which surface salinity might not). But actually the one I think is really wrong is whether deep water is more saline than surface water – given surface salinities of up to 37 psu, the deep ocean values of 34 to 35 are not higher. – gavin]

  16. 16
    Adam says:

    Didn’t read the other articles, but just for the record, the BBC article discusses salinity, despite the headline.

  17. 17
    David Beach says:

    The Aquarius passive microwave measurements are exclusively sea surface related. Is this what is wanted? I would have thought that bulk (deeper) measurements were needed for ocean current effects to be quantified, but I guess the global coverage outweighs that. Comments?

    [Response: Definitely. But there are limits to what can be remotely sensed. The sea surface changes themselves will be a big increase in info compared to today since we don’t have any ‘synoptic’-scale observations at the global scale right now. There are a few continuous arrays – TOGA, 24N, ocean stations etc. but these don’t have the spatial coverage needed. – gavin]

  18. 18
    Terry says:

    10/10, not bad for a physical chemist. The only ambiguous qn was the one about future prediction, given that an increase in ice melt indicates a decrease, but being the skeptic I am I chose “dont know”.

  19. 19
    sam marshall says:

    8/10. Yeah, nothing really to contribute, just ‘woohoo I got more points than the actual scientist’. :) Although, at least a couple of those I wouldn’t have got without reading this just prior, which is a slightly unfair example.

    I got the one about whether it varies a lot wrong; had I actually paid attention to the scale on the chart in this article, I would have agreed with them that it ‘doesn’t vary much’. In lay terms that seems about right. If you’re in a car going 30, and somebody else on the same road is going 35, is that varying a lot or not much? I’d say the latter. (Presumably in this case the differences are significant but that doesn’t mean they are large.)

    Thanks for the article.

  20. 20
    Jeff L. says:

    I also guessed that deep water would have lower salinity, because I was thinking of the all the cold fresh water created by melting ice sheets in the polar regions.

    [Response: Well, see my response above. North Atlantic Deep Water is around 34.9 psu – nothing like as salty as the tropical surface waters in the Atlantic (around 37 psu) – but it is substantially cooler. Antarctic deep water is a colder and a little fresher – but of comparable density. So the question is rather ill-posed. – gavin]

  21. 21
    Terry says:

    All things being equal, thermodynamics says that salinity will increase (only a weak dependence) with increasing depth, but as Gavin points out the tropical waters are saltier at the surface to begin with, by more than the pressure dependence, so actually the answer to the question depends on which ocean you are referring to.

  22. 22
    Thomas says:

    Sea Level rise on the order of centimeters (or even a meter or two) is pretty small compared to the average depth of the ocean (roughly 4KM), so it would only dilute things (if well mixed) by less than a part in ten to the fourth (or less). Maybe thats why they considered it not to matter, since that secular change is much smaller than the spatial variations.

  23. 23
    Greg Simpson says:

    I missed the questions about whether the deep water is more saline, the question about whether the ocean will be less saline in the future, and, despite staring at the graph here for minutes, what ocean is the saltiest. I knew the graph showed the Atlantic as the most red, and I knew that red meant saltier, but I still confidently chose the Pacific!

    So, 7 of 10.

  24. 24
    Jean-Pierre Gattuso says:

    Since 1978, salinity is measured on the Practical Salinity Scale 1978 (Unesco, 1981). It is defined as the ratio of two conductivities and therefore has no unit and PSU should not be used (Ridout, 1997).

    Ridout, P.: Salinity units, Ocean Challenge, 7(1), 4, 1997.

    Unesco: Background papers and supporting data on the Practical Salinity Scale 1978, Unesco Technical Papers in Marine Science, 37, 1-144, 1981.

    [Response: The new TEOS-10 ‘Absolute Salinity’ scales and recommendations are likely to replace these older standards over time – see for instance. – gavin]

  25. 25
    Magnus W says:

    Am I missing something?

    Question 9, Climate change is adding more salt to the ocean…

    A warmer world = more weathering = more salt in the ocean?


    [Response: The weathering/burial feedbacks are really long-term (~100,000 years), so they have to be small on the kind of timescales we mostly care about. The changes over decades will be related to a) land and sea ice melt (lowering salinity), b) increased hydrological cycle (more evap and more precip, thus increased surface gradients), c) groundwater changes (dams, aquifer depletion, etc.). Ocean circulation changes are probably second order. All of these will be much bigger than weathering changes. – gavin]

  26. 26
    RussH says:

    I got 8/10, I thought deep water would be more saline and also the oceans would be less saline due to climate change.

  27. 27
    JimCA says:

    9 out of 10 because I guessed Indian over Atlantic, assuming that the Indian had more evaporation and less river influx. I’d guess that question does the best job of discriminating between climate experts and lay people such as me.

    Regarding the question about salinity at depth:

    The question asks if salinity is higher at the sea bottom than at the surface. I think the only plausible reading is to view that as a comparison within one vertical column. Alternatively you could sum over all columns and get global averages, but that amounts to essentially the same thing. Since higher salinity leads to increasing density, everything else being equal (and for global averages that’s usually the best assumption) deeper should be saltier.

    The alternative would require a global predominance of warmer saltier water overlying colder fresher water, which is conceivable but seems intuitively unlikely.

  28. 28
    Some European says:

    Heehee! Everybody’s fond of the quiz! I got 7/10 myself, therefore I’m just as expert in ocean salinity as Gavin Schmidt!
    The lesson to be learned: quizzes are a cool tool to educate people. Maybe Climate Progress and friends should post some quizzes to attract attention. And maybe I should post one on my facebook profile. I got on facebook for the sole reason of reaching out to my friends and family about the climate crisis. But no one seems to be paying attention to my numerous links.

    On media coverage. The Belgian article that I read had this awful headline, framing the issue in such a way that anti-science comments were unavoidable: “Research into the saltiness of seawater … from space”. You can imagine what people think: “in stead of just taking samples with a bucket, these ‘scientists’ are wasting our tax money with fancy satellites”

    Note to researchers: don’t send the preliminary data analysis to the Daily Mail, to avoid any “Ice age is near!” headlines followed by “Ice age scare hyped by environmentalists now debunked by satellite measurements” a couple of years later.

  29. 29
    One Anonymous Bloke says:

    7/10. On a wider note, it’s good to see a useful satellite making it past institutional hurdles and random failure into orbit.

  30. 30
    Peter says:

    7/10 for this biologist, but I’m feeling cheated now as according to Gavin’s response to #15, I should have got 9/10!

    On the last question, I suppose it depends how sure you have to be before you can say you “know” what future salinity changes will be, but I figured (simplistically) that greater evaporation and rainfall in a warmer climate wouldn’t necessarily change the proportion of rain falling on land or sea, or alter the fact that rain falling on the land makes its way to the sea sooner or later.

    On the relative salinity of surface and deep water, I thought that there was little mixing of the two, and therefore that evaporation would mainly affect the surface layers.

    Oh, and I second the comments about “pounds” of seawater. And degrees Fahrenheit all over the linked climate change info pages. I thought NASA had gone metric?

  31. 31
    ldavidcooke says:

    Hey Dr. Schmidt and Company,

    I think this may be OT and for that my apoligies. However, I am just a bit curious, if the Odin sensor set included a sodium capable detector and it could be reoriented to the surface rather then the Mesosphere then why launch Aquarius? As I understand it Aquarius has no sodium photon emission spectrum detector. Also to the best of the released information Aquarius is primarily a high resolution thermograph detector meaning it is mainly a high resolution (93km) thermometer.
    I know I may be ignorant; but, how does a thermometer read the sodium / maganese / chloride,…etc or possibly sulfuric/carbonic acid densities in the “boundary layer” when there are no less then 30 different degrees of freedom of varibility of temperature variation? Apparently my lack of scienctific/math education restricts my ability to extract the sodium signal from a 1 ghz microwave scanner thermal signal… Maybe a simple explanation from the experts here will help.

    Dave Cooke

  32. 32
    Sekerob says:

    Just wondered on the amount of salt that 7 billion humans use and insert into the environment… not looked for what ends ‘net’ into the oceans:

    Data per 2006

    2011 Outlook

    Solar evaporation of seawater or inland brines accounts for 40% of salt production, the remainder is divided between mined rock salt and solution mining.

    Forecast use 300 million Mt by 2015.

    In the Sahara and other places land is irrigated by water from aquifers, to the point that the piece of land eventually has taken on so much natrate that it becomes unusable. It’s an issue as recently presented in a program for countries as The Netherlands with substantial agricultural produce also in context to SLR [forecasts: more rising]. Not sure what happens with salt from seawater desalination plants. If dumped back I’d suppose there to be localized disturbances to the ecosystem.

    Added: Yes, it’s dumped back

  33. 33
    Magnus W says:


    It might be my English but the question is not of the concentration?

    Yes, it would be small… I have not tried to calculate it just remembered some research done on Island (I think) that already shows increased weathering mostly due hotter temperature if I remember correctly.

  34. 34
    Mark says:

    I psyched out the questions, guessing what they were after, rather than straight up answers,so I got 9/10, but I wouldn’t have made the Atlantic the saltiest w/o the map in the post. I would have missed the variation question if I answered it as a physical scientist, but I guessed the quiz is designed for more general audiences and said the salinity variation is small. I missed the deep versus shallow salinity. North Atlantic Deep and Antarctic Bottom Waters are more saline than intermediate waters, but also very cold, so some very warm surface waters can be saltier and still ‘float’. I would offer that scores in the 6/10 to 8/10 range contain two groups, one non-oceanographer group that knows quite a bit for a general audience quiz, and an oceanographer group that knows ‘too much’ for a general audience quiz.

  35. 35
    Jeremy says:

    Peter in #30 “I thought NASA had gone metric?”

    Yeah, how many Mars lander impact tests do they need to make?

  36. 36
    Walter Pearce says:

    6 out of 10 (red faced). Love this type of quiz, but even more — reading the deeper discussion of the questions. Thanks!

  37. 37

    10/10 for me, albeit with an asterisk, since I’d already read a couple of the comments. (Actually, they prompted me to take the quiz, thinking “Better do it before I come across ALL the answers!”)

    The one I was a little proud of was the last, the ‘what will happen to salinity’ question. Thought about the melting ice adding freshwater, the increased evaporation on the other hand. Thought about the intensified hydrological cycle–presumably evaporation and precipitation must balance over time so that’d be a wash (yuk, yuk!), whereas increased erosion might mean increasing salt transport, too.

    Then I thought about trying to constrain all that quantitatively, and went for ‘don’t know,’ which proved correct.

  38. 38
    Sekerob says:

    Question 11: What evaporates easier… Sweet or Saline Water?

    Answer 11: Sweet

  39. 39
    Manu says:


    The 1.4 GHz of sea water is far from pure thermal (i.e. black body), it is modified by the dielectric properties of sea water. The salt in the water change the ionic conductivity, and therefore the natural emission. That is how the change in salinity is being measured by the radiometers (after correction of other concurrent effects on emissivity, e.g. changes in temperature).

  40. 40
    ldavidcooke says:

    Re: 39

    Hey Manu,

    Yes, I went on to read a bit more after asking the question. The works of reference wrt Radiative Transfer Method provided by Dr. Schmidt and a few papers by Dr. Germer at the NIST and later Dr. Mulholland helped explain the process/procedure. You are correct that it appears at 1.4 ghz using the Klien-Swift model the reflectivity of the oceans surface as a function, including temperature and surface roughness should provide fairly good data via polarization measurements of scattered signal collection. The primary “fly in the ointment”‘s appearing to be accurate thermal “blackbody” SST references and sea surface variation/smoothness.

    My initial concern was for a lack of direct measurement of the mix and quantity of ionic substances in the water; however, the models appear to offer a rough or 90% est. of the values of NaCl that should be in the “boundary” layer. The only major concern I have remaining is the sponsoring organization insuring comparable “real world” sampling for the purpose of validating the model baselines is carried out asap. With the purpose being to reduce the injection of error or adding fuel for questions by folks unfamiliar with the dielectric or reflective/refractive function of sea water on low frequency microwave energy.

    In the meantime, has anyone an insight of the anticipated error bars, as a function of salinity, based on the current design. I had not seen any “specs” rolled out in layman language, though a manufacturer of a similar measurement device has suggested a 20 fold variation just due to surface variation…

    Dave Cooke

  41. 41
    Manu says:


    Obviously, I can’t tell you about actual error bars yet, but the objective of the mission is an accuracy of 0.2 psu on monthly maps (at about 150 km spatial resolution). The globe is mapped entirely every seven days, then these seven-day maps are averaged together in order to (hopefully) bring down the random error for the resulting monthly maps.

    Now the actual error is going to depend on a few parameters: actual accuracy & precision of the instrument (this should be pretty good) and of the geophysical “corrections” to apply to retrieve salinity (I have already mentioned the sea surface temperature). Among the geophysical corrections, the most critical and problematic is probably going to be surface roughness (i.e. wind & waves at surface to keep it simple), but there are others that are usually not so problematic, but could be here because of the stringent accuracy required for salinity retrieval (about 0.05 K on the brightness temperature, this is very challenging !). However, Aquarius has a great advantage here compared to SMOS: it embarks a scatterometer (i.e. radar) that should provide essential info about surface roughness at the time and location of the radiometer measurements.

  42. 42
    ldavidcooke says:

    Hey Manu,

    Thanks, that is pretty aggressive, a 95% accuracy across 150km is outstanding in a remote sensing system.

    (Even the gross level CO2 measurements I have seen from the Japanese “Bird” looked alot like water vapor scattering as opposed to CO2 plumes. (Then again if the measures are made during the “day cycle”, the reflected energy can overdrive the sensor, even with the polarizer and spectral filters in place…))

    Again my thanks for your timely and informed response.

    Dave Cooke

  43. 43
    Garry S-J says:

    Re question 5 – since when was “chloride” an element?

  44. 44
    Garry S-J says:

    Also, re question 10 – How can I be incorrect if I answer that climate change will likely make the oceans less salty?
    I mean, if the answer is that scientists don’t know, then how can they know I’m incorrect? Maybe the question should begin “what do scientists expect…” or the like.

  45. 45
    John E. Pearson says:

    8/10 . I missed the one that asked whether salinity is small or not. I answered “NO” because I was thinking of the large fluxes that are brought about by salinity. When you say something is “small” one normally has to specify what it is small relative to. Ambiguous question. I also got the last one wrong and said that salinity should decrease thinking that if sea level went up because of melted ice that salinity would drop.

  46. 46

    #44–Well, the quiz writer was writing out of authority (misplaced or otherwise, I can’t say.) And the logic makes sense: if you take the quiz, you tacitly accept the authority of the quiz maker, in the sense that you are willing to posit the state of their knowledge as more developed than yours. And if you posit that, then if they don’t know, how do *you* know?

    Of course, that grant of authority can be withdrawn at any time–Gavin’s questioning of the soundness of the salinity question is an example. So is much climate skepticism (with or without quotes)–which makes the further point that the legitimacy of the authority and its challenge alike should be judged on a case by case basis. “They laughed at Galileo, but they also laughed at Bozo the clown.”

  47. 47
    spyder says:

    Just curious if this satellite data will make sense of the thermoclines across the depths, that must play a role in salinity of the water?

  48. 48
    Maya says:

    9/10. I would have gotten them all right if I’d looked back at the graphic, but that seemed like cheating. So I guessed the Arctic was the saltiest, since it’s the coldest, but apparently evaporation trumps density.

  49. 49
    Gerry Quinn says:

    7/10, but I think two of the ones I got wrong, I was actually right. Water at the seafloor must obviously be less salty on average because evaporation takes place at the surface. And if melting ice is causes sea levels to rise over the next century, with no obvious reason for extra salt to be added, the oceans should grow less salty.

    I guessed the Indian ocean was saltiest because I figured it was yje hottest and should evaporate more.

    I got one question right by their standards but morally wrong – if I had thought the salt content of the oceans varied by as much 32-37, I’d have said the variation was large.

  50. 50
    Joseph Sobry says:

    8/10 . What if rising sea levels invade a previously dried out salt deposit ? This has happened before according to some stuff I have read. The mediterranean comes to mind.
    Of course there may be no salt deposits within reach of sea level rise in the next 100 years.