The Earth generates a global magnetic field—the geomagnetic field—from the activity in its iron core. This field can be used to show where magnetic rocks in the crust distort it. That's something like the way we use X-rays to map your teeth or bones based on their effect on the rays.
Large-scale magnetic surveys are generally done using aircraft, which fly a magnetometer over an area in a carefully controlled grid at a specified altitude. The resulting profiles of magnetic data are integrated into aeromagnetic maps. What might we learn from these?
First we need to know some things about what magnetic data might show. The most important thing is that rocks cannot be magnetized unless they're cool. More precisely, their minerals have to be below their Curie temperatures (see "How Are Rocks Magnetized?" for more). The foremost mineral to consider for aeromagnetic work is magnetite (pyrrhotite and titanium-bearing hematite are the others), and its Curie temperature is 680°C. That temperature corresponds roughly to about 30 kilometers in continental rocks (it varies a lot), so we're talking about the crust or upper lithosphere. Everything beneath that, all the way down to the core itself, is magnetically transparent.
The next thing to know is what rocks are magnetic. Magnetic petrology is a complex field. But in general, sedimentary rocks are non-magnetic. This means that aeromagnetic surveys can see right through them, which can be extremely useful.
Among igneous and metamorphic rocks, generalizations are less useful. Granites are low in iron, which concentrates in the dark minerals as well as in primary magnetite, but not always. Gabbros and especially ultramafic rocks tend to show up in magnetic surveys, but not always. Serpentinite can be particularly rich in magnetite because of the chemical process that forms it, but not always. Bodies of ore, large or small, may have strong aeromagnetic signals, but not always. Faults and intrusions may be obvious—or invisible. It's not like regular geology, but aeromagnetic information can be very valuable.
Aeromagnetic interpretation is a demanding discipline that relies on digital skills as much as visual ones. What's meaningful is not obvious to the eye. The task is like trying to make a drawing of a large, irregular sculpture of clear glass, with only its shadow on a wall to work from. A particular blob can be the product of a wide range of possibilities giving the same result. This means that every other source of information—geologic maps, seismic profiles, gravity mapping, magnetotelluric data, drilling and mine data and more—needs to be considered to get the most from aeromagnetic data.
Aeromagnetic features are "illuminated" by a geomagnetic field that's slanted almost everywhere on Earth. Magnetic field lines are vertical only near the magnetic poles; everywhere else, magnetic inclination offsets aeromagnetic features from their actual positions. And whereas an aerial photo shows the Earth's surface with great accuracy, aeromagnetic maps include features that may sit at various distances, from the ground surface to the base of the crust. This makes their images out of focus, with no way to focus them except by making assumptions and best guesses.
Interpreters of aeromagnetic data proceed by first removing as much distortion as geometry safely allows. Then they start matching the shapes of features to models that are based on geometric shapes, like wedges and slabs and cylinders, that correspond to typical geological features. This is not intuitive. A fault or a lava intrusion (a dike or sill) may look like one only if it exactly aligns with the magnetic field, with the right direction and the right tilt—and only if it contains magnetic minerals. Otherwise it may look like a strange set of blobs, or even nothing at all.
They can also make note of patterns, like aligned features, that correspond to the large-scale fabric of the mapped area. The aeromagnetic map of a mountain range, for instance, can be expected to show features running in the same direction as the mountains. Areas with lots of plutons should have round aeromagnetic outlines that match them.
In skilled hands, aeromagnetic mapping can uncover deeply buried or chemically obscure features that shed light on an area's tectonic history, its mineral resource potential and its hidden structure. Conversely, the absence of aeromagnetic features can be a handy sign of sedimentary rocks in thicknesses suitable for oil exploration. It's all in the way you look at the data.