Satellite Observations of Magnetic Fields Due to Ocean Tidal Flow

See allHide authors and affiliations

Science  10 Jan 2003:
Vol. 299, Issue 5604, pp. 239-241
DOI: 10.1126/science.1078074


The ocean is an electrically conducting fluid that generates secondary magnetic fields as it flows through Earth's main magnetic field. Extracting ocean flow signals from remote observations has become possible with the current generation of satellites measuring Earth's magnetic field. Here, we consider the magnetic fields generated by the ocean lunar semidiurnal (M2) tide and demonstrate that magnetic fields of oceanic origin can be clearly identified in satellite observations.

In a fully magnetohydrodynamic process, the flow and electromagnetic fields are coupled. In the ocean, however, flow generates electromagnetic fields but the electromagnetic fields are not thought to affect the flow appreciably. This reduced magnetohydrodynamic case is often called “motional induction” and can be understood as follows. The dissolved salts in seawater form hydrated, electrically charged ions. As the charged ions are carried by the ocean flow through Earth's main magnetic field, they are deflected by the Lorentz force, which acts in a direction perpendicular to both the velocity and magnetic field. This leads to various combinations of two effects. First, the migrating ions can accumulate to form electrical spatial charge densities that in turn create electric fields that tend to prevent further migration of charge. Second, the spatial charge densities can be relieved by electrical shorting through surrounding sections of the water or electrically conducting sediments. The latter effect involves electrical currents and the associated secondary magnetic fields, which are the subject of this paper.

Two components of the ocean-generated magnetic field can be distinguished. The first is a “toroidal” component that has been estimated to reach maximum amplitudes of 100 nT but is confined to the ocean and sediments and is therefore not observable remotely (1–5). This component results from electric current circuits closing in planes containing the vertical axis. The second is a much weaker (1 to 10 nT) “poloidal” component with large spatial decay scales that allow the magnetic fields to reach remote land and satellite locations (4,6–10). This component involves electric current circuits closing horizontally and is the least understood because it is generated by large-scale integrals of ocean flow transport and estimates typically require large-domain integrations.

But this dependence of the far-reaching poloidal magnetic fields on transport integrals also makes these fields attractive. In principle, information about past and present ocean variability is contained in the land and satellite magnetic records, and this variability would primarily reflect integrated transport quantities (including in ice-covered regions) that are difficult to obtain using other methods (11). Understanding such ocean variability is a key factor in addressing climate and global change concerns, and although an assessment of the potential for exploiting the magnetic fields in this way is beyond the scope of this paper, here we describe an initial step in identifying ocean flow effects in the satellite record. We have performed a global numerical prediction of the magnetic fields due to the semidiurnal M2 ocean tidal constituent and have compared it with observations made aboard the CHAMP (Challenging Minisatellite Payload) satellite, and we found close agreement between the observations and predictions.

In the numerical prediction, the tidal flow and main magnetic field used are given by the results (TPXO.5.1) of Egbert (12) and the CO2 model (13), respectively. The model formulation is based on a thin-sheet induction equation [similar to the formulation discussed first by Price (14) but with several modifications made for the case of ocean flow forcing (15)] coupled to equations describing zero-laplacian magnetic potentials outside the shell (16). Independent of the prediction, the periodic M2magnetic-field variations were extracted from CHAMP satellite measurements collected over 2 years (16). Figure 1 displays their real and imaginary parts against the corresponding model predictions (see also movie S1). A comparison of their spectra over ocean and land (Fig. 2) confirms that the observed M2 signal is stronger over the ocean.

Figure 1

Predicted and observed magnetic signal (scalar anomaly) of the M2 ocean tide. Along-track filtering to remove large-scale (>10,000 km) magnetospheric fields has been applied to both. Representing the periodic signal in terms of its real component Mr(phase = 0°) and its imaginary component Mi (phase = 90°), the observations are displayed in (A) and (C) against the corresponding predictions in (B) and (D) (see also movie S1). The rectangles in (A) and (C) indicate the areas from which the spectra of Fig. 2 were computed.

Figure 2

Magnetic intensity spectra of the data from the areas indicated in Fig. 1. Although the ocean area exhibits a clear M2 peak, this is absent from the land data.

The predicted and observed ocean-generated magnetic fields agree remarkably well. The long bands of enhanced amplitude (yellow and blue stripes in Fig. 1) running across all longitudes are very much the same in the observation and the model results. A closer inspection reveals that even regional peaks (red or dark blue spots) can be found to match in most cases. Meridional stripes, visible in Fig. 1C for the Pacific Ocean, are probably a line-leveling problem due to incompletely removed magnetospheric fields. An assimilation of the data into the model could remove this effect, but we regarded a completely independent treatment of measurements and model as important for this first test. In rare cases, weak anomaly centers appear unrealistically over the source-free land areas. This is an effect of the along-track filtering (applied to both model and observed data), which transports signal from the ocean onto the land areas. Another effect of filtering is to reduce the local amplitude peaks of the predicted signal, which at satellite altitude reached 3 nT before filtering. Both of these filtering effects are illustrated in fig. S1.

As a check for potential systematic differences in the amplitudes, Fig. 3 shows the amplitudes of the observed and the predicted signals averaged over latitude and plotted against longitude. Apart from the aforementioned line-leveling problem (the short-scale oscillations with longitude), the curves track each other closely both in shape and amplitude despite the fact that the predictions were calculated assuming an insulating mantle. This model assumption was initially made to provide an upper-bound estimate while avoiding somewhat ad hoc assumptions regarding the mantle conductivity structure. We anticipated that the predictions would systematically overestimate the longer wavelengths due to the absence of coupling with the lower mantle. Though mantle effects at some level are both expected and possibly evident in our results, a discussion of this does not appear to be required for the primary purposes of this paper and is postponed.

Figure 3

Meridionally averaged magnetic signal amplitude. Predicted and observed amplitudes are almost congruent. In two regions there are, however, obvious deviations. Around –90° longitude, the model underestimates the tidal signal. This longitude maps to the Gulf of Mexico but also to the west coast of South America. Conversely, the model predicts too large a signal in the Indian Ocean (90° longitude).

Focusing on the lunar M2 tide, it has been shown here for the first time that the ocean flow makes a substantial contribution to the geomagnetic field at satellite altitude. This has important implications: In broader terms, it encourages future studies to assess the feasibility of monitoring ocean flow from space. A more immediate consequence, however, is that it shows that oceanic signals must be incorporated into geomagnetic field models. Indeed, with recent advances in internal and external field separation (17), the ocean flow signal is now the strongest remaining signal in the low-latitude magnetic residuals that has not yet been modeled. Correcting magnetic readings for ocean flow signals could strongly improve the accuracy of lithospheric anomaly maps and greatly raise the detectability of small-scale crustal magnetization.

Supporting Online Material

Materials and Methods

Fig. S1

Movie S1

  • * To whom correspondence should be addressed. E-mail: tyler{at}


View Abstract

Stay Connected to Science

Navigate This Article