Bob Grimm - EM Sounding on Mars

ELECTROMAGNETIC SOUNDING FOR GROUNDWATER ON MARS

See Grimm, 2002 for a comprehensive review

Detection of liquid water on Mars is a key exploration objective, but geophysical methods currently in use or under consideration for subsurface exploration are not optimal for water detection.

Subsurface-sounding radars (aka GPR) are currently operating in Mars orbit MARSIS and SHARAD). While subsurface echoes have been detected, there has been no indication of subsurface water. The high dielectric constant of water is commonly cited as diagnostic but the rock porosities generally restrict natural dielectric constants to less than 10, even for wet materials. It's very difficult for orbital radars to measure dielectric constant directly so it is assumed that some high-contrast interface must represent a water table. Because radar responds to all dielectric contrasts, it's really a technique that is best suited for mapping structure and stratigraphy. Depths of investigation in rock have also been overestimated by proponents who have neglected scattering and absorption from small quantities of interfacial water at subfreezing temperatures. Radar is still very useful in planetary subsurface sounding, so it's not neglected in my current research.

Seismology is an important tool for understanding the interior of Mars, but not for finding water. Active seismic surveys are impractical compared to radar. Passive measurements (using signals from distant earthquakes) can assess pore compressibility from P/S velocity ratios, but this technique is insensitive above a few tens of MPa confining pressure, i.e., at depths greater than a few kilometers on Mars. The cryosphere of Mars, on average, is likely thicker than this, so stations would have to be fortuitiously positioned over special regions of shallow (say <1 km) groundwater.

Low-frequency electromagnetism is well suited to identifying and characterizing groundwater on Mars.

At lower frequencies (<1 MHz on Earth, <1 kHz on Mars), EM in the ground is inductive, and energy moves by diffusion instead of wave propagation. Low-frequency EM penetrates more deeply and is less affected by small-scale heterogeneity than radar. It is sensitive to electrical conductivity instead of dielectric constant; contrasts in the former range over many orders of magnitude, providing strong discriminants. On Mars, groundwater that has not been in meteoric circulation for aeons (if ever) will likely be saline and therefore will be a near-ideal, conductive EM target. Several low-frequency EM techniques are applicable to Mars:

Magnetotellurics (MT). This is a classic, passive method that uses natural EM signals for subsurface sounding. The horizontal components of the electric and magnetic fields are used to compute ground impedance and hence resistivity. The depth of investigation increases with wave period according to the skin depth effect and standard inversion procedures yield a depth-conductivity profile.


MT sounding with Mars-prototype system (G. Delory, PI) on the Eastern Snake River Plain (ESRP), Idaho. ESRP is the terrestrial type location for planetary plains-style basaltic volcanism but is underlain by aqueously altered tuffs. This contact is evident in the inversion (right panel) at ~800-m depth, in agreement with regional well control. Electric and magnetic field time series have been fourier transformed and converted to apparent resistivity (upper left panel) and phase (lower left panel) for inversion input.

It is often said that geophysical inversions are nonunique, but this is an oversimplification. All inversions, right down to a least-squares straight-line fit, have uncertainties in the solved parameters. In MT, this is can be manifested as simultaneous uncertainty in the thickness and conductivity of a layer. Yet the product of these parameters (the conductance) is often determined accurately. Furthermore, the most fundamental parameter to be assessed for Mars - the depth to groundwater - will be very robustly determined because of the large conductivity contrast and relatively simple geometry (conductive aquifer under resistive cryosphere).

Natural signals used by terrestrial MT include daily ionospheric currents (Sq), magnetospheric pulsations, and lightning. We don't yet know exactly what the natural electromagnetic spectrum looks like near the surface of Mars. It has an ionosphere that interacts strongly with the solar wind; it lacks a global magnetosphere but has "mini-magnetospheres" due to intense crustal magnetization. Lightning has not yet been detected, but is suspected due to the abundance of mobile, electrically chargeable dust. Any ambient energy is useful for MT sounding!

Geomagnetic Depth Sounding (GDS). This variant of MT exploits the same natural signals, but only measures magnetic fields. Now, however, the vertical component and the horizontal gradient of the horizontal component must be included to complete the impedance calculation. GDS has traditionally used magnetometer arrays to measure these horizontal gradients (~10 km station spacing, ~100 km array size, two dimensions). The ESA Mars NetLander mission was to implement a minimal GDS array with 4 stations, but this would have strongly undersampled horizontal gradients at most frequencies. However, single or widely separated stations on Mars can still estimate the dominant spatial frequency (and hence horizontal gradient) for ultralow frequency signals tied to planetary rotation: Sq and its harmonics. Because the computed resistivity varies weakly with the source field at long wavelength, this has little affect on the ability to detect and characterize groundwater.

Time-Domain Electromagnetic Method (TDEM or TEM). This is an active EM method that supplies its owns source field: there is a transmitter as well as a receiver. Therefore this method can be considered as a backup for Mars in the event of extraordinarily weak natural fields or to fill specific frequency gaps. TDEM is widely used to measure ground conductivity as a function of depth. The most common implementation uses a large transmitter loop (~100 m in diameter) and a compact, multiturn receiver loop. When electrical current is abruptly extinguished in the transmitter loop, eddy currents flow in the ground, depending on conductivity, and their magnetic fields are detected at the receiver. TDEM does not depend on geometrical detail (like a radar antenna) of the coils yet is very sensitive to subsurface conductors.

Surface Nuclear Magnetic Resonance (SNMR). This method, still new to terrestrial geophysics, uses a time-varying applied field to tip spinning hydrogen nuclei away from the planet's static magnetic field, and the synchronous rotation of magnetic moments is detected in a receiver coil. Because "free" hydrogen nuclei have a characteristic Larmor frequency, this is the only subsurface sounding method that uniquely detects water. The setup with transmitting and receiving loops is similar to TDEM, but greater power is necessary. Application to Mars was reviewed by Grimm, 2003.

Summary. A variety of geophysical methods can be applied to the search for groundwater on Mars. Low-frequency electromagnetic methods are favored. Magnetotellurics in particular is the best trade of resources required versus information gained.

Geophysical Methods to Detect and Characterize Groundwater on Mars

Technique

Depth of Investigation

Resources

Mobility?

Sensitivity to Water

Radar

few km

Med

Yes

Medium

Seismic

few km

Low

Maybe

Low

Magnetotellurics

many km

Low-Med

Yes

High

Geomag. Depth Sounding.

many km

Low

Yes

High

Time-Domain EM

few km

High

No

High

Surface Nuclear Magn. Reson.

< 1 km

High

No

Highest (unique)