I have a question about a module on Stommel's Two-Box model.

I am uncertain about the following density equation: $ \Delta \rho = | R \Delta S - \Delta T| $

A larger difference between temperatures/salinity levels across the two regions of the North Atlantic should both increase the density difference. Why is this a subtraction? Is this because of the opposite relationship between salinity/temperature at each pole (i.e., high salinity increasing the density at the equatorial region with high temp decreasing density at the equatorial region, and vice versa at the poles)?

  • $\begingroup$ Have you looked at the paragraph "It is important to understand an important negative feedback in this system involving Q and the pole-equator temperature difference ΔT. When the temperature difference, ΔT, is great, then Q will be strong, but this transports more heat to the polar region, thus lessening ΔT, which in turn weakens Q. On the other hand, if Q is weak, then there is little heat transport to the polar region, allowing this region to grow colder, thus increasing ΔT, which can then in turn leads to a stronger Q and thus more heat to the polar region." on the page? $\endgroup$ Commented Feb 3 at 18:25

1 Answer 1


Modifiing my previous answer thanks to you comment, I realized I not clear at all. Here it's better.

First, I want to state that the notations in your material seem weird to me as the normalizing and adimentionnality of the values are not properly presented. For the eq. 1 and 2, $T_{eq}$ and $S_{eq}$ are said to be dimensionless yet they are added to $\Delta T$ and $\Delta S$ that are not introduced as dimensionless.

In any case, it seems akin to the density anomaly (see THIS LINK, page 6). The same pdf file the link leads to bears a thorough explanation of the model (page 9 and following).

Without going trough the nondimensionalization process, by taking the density equation of salty waters, $\rho=\rho_0(1-\alpha T+\beta S)$ (increases with increasing S and decreases with increasing T with rate respectively $\alpha$ and $\beta$), you have that \begin{align}\Delta \rho&=\rho_{A}-\rho_{B}\\&=\rho_0(1-\alpha T_{A}+\beta S_{A})-\rho_0(1-\alpha T_{B}+\beta S_{B})\\&=\rho_0(-\alpha \Delta T+\beta \Delta S).\end{align} Using $R=\frac{\beta}{\alpha}$ and $\Delta \rho_*=\frac{\Delta \rho}{\alpha\rho_0}$ leads to $$\Delta\rho_*=-\Delta T+R\Delta S.$$ I'm not sure about why the absolute value is introduced, maybe because the textbook authors weren't interested in wich part of the ocean was the most dense?

You can take a look at the original paper by Stommel at he following LINK.

  • $\begingroup$ A higher temperature difference across two regions of the ocean also means a higher density difference. The detlaT and deltaS are between 0 and 1 (always positive): T_equatorial-T_polar = Positive value; S_equatorial - S_polar = Positive value. $\endgroup$
    – AAAA
    Commented Feb 3 at 18:13
  • $\begingroup$ Hi @AAAA , I modified my answer to be more thorough . I'm not absolutely sure of being right as it's been a while since I've studied this model. However, your base material is perplexity inducing as it jumps over large parts of the explanation and seems to neglect some precisions - I recommand you to look at other material for a finer understanding. $\endgroup$
    – Varazak
    Commented Feb 4 at 23:15
  • $\begingroup$ Great! Thanks. Yes, I have been using my mathematics texts but was hoping to see if there was a more intuitive way of understanding this model. Thank you for the explanation! I will keep working on it to figure out how else it can be presented. $\endgroup$
    – AAAA
    Commented Feb 6 at 22:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.