In geophysics, a magnetic anomaly is a local variation in the Earth's magnetic field resulting from variations in the chemistry or magnetism of the rocks. Mapping of variation over an area is valuable in detecting structures obscured by overlying material. The magnetic variation in successive bands of ocean floor parallel with mid-ocean ridges is important evidence supporting the theory of seafloor spreading, central to plate tectonics. In geophysics, a magnetic anomaly is a local variation in the Earth's magnetic field resulting from variations in the chemistry or magnetism of the rocks. Mapping of variation over an area is valuable in detecting structures obscured by overlying material. The magnetic variation in successive bands of ocean floor parallel with mid-ocean ridges is important evidence supporting the theory of seafloor spreading, central to plate tectonics. Magnetic anomalies are generally a small fraction of the magnetic field. The total field ranges from 25,000 to 65,000 nanoteslas (nT). To measure anomalies, magnetometers need a sensitivity of 10 nT or less. There are three main types of magnetometer used to measure magnetic anomalies::162–164:77–79 In ground-based surveys, measurements are made at a series of stations, typically 15 to 60 m apart. Usually a proton precession magnetometer is used and it is often mounted on a pole. Raising the magnetometer reduces the influence of small ferrous objects that were discarded by humans. To further reduce unwanted signals, they do not carry objects such as keys, knives or compasses. In addition, objects such as motor vehicles, railway lines, and barbed wire fences are avoided. If some such contaminant is overlooked, it often shows up as a sharp spike in the anomaly, so such features are treated with suspicion. The main application for ground-based surveys is detailed search for minerals.:163:83–84 Airborne magnetic surveys are often used in oil surveys to provide preliminary information for seismic surveys. In some countries such as Canada, government agencies have made systematic surveys of large areas. The survey generally involves making a series of parallel runs at a constant height and intervals of anywhere from a hundred meters to several kilometers. These are crossed by occasional tie lines, perpendicular to the main survey, to check for errors. The plane is a source of magnetism, so sensors are either mounted on a boom (as in the figure) or towed behind on a cable. Aeromagnetic surveys have a lower spatial resolution of ground surveys, but this can be an advantage for a regional survey of deeper rocks.:166:81–83 In shipborne surveys, a magnetometer is towed a few hundred meters behind a ship in a device called a fish. The sensor is kept at a constant depth of about 15 m. Otherwise, the procedure is similar to that used in aeromagnetic surveys.:167:83 The Sputnik 3 spacecraft in 1958 was the first to carry a magnetometer.:155 In the fall of 1979, Magsat was launched and jointly operated by NASA and USGS until the spring of 1980. It had a caesium vapor scalar magnetometer and a fluxgate vector magnetometer. CHAMP, a German satellite, made precise gravity and magnetic measurements from 2001 to 2010. A Danish satellite, Ørsted, was launched in 1999 and is still in operation, while the Swarm mission of the European Space Agency involves a 'constellation' of three satellites that were launched in November, 2013. There are two main corrections that are needed for magnetic measurements. The first is remove short-term variations in the field from external sources. There are diurnal variations that have a period of 24 hours and magnitudes of up to 30 nT, probably from the action of the solar wind on the ionosphere.:72 In addition, magnetic storms can have peak magnitudes of 1000 nT and can last for several days. Their contribution can be measured by returning to a base station repeatedly or by having another magnetometer that periodically measures the field at a fixed location.:167 The anomaly is the local contribution to the magnetic field, so the main geomagnetic field must be subtracted from it. Usually the International Geomagnetic Reference Field is used for this purpose. This is a large-scale, time-averaged mathematical model of the Earth's field based on measurements from satellites, magnetic observatories and other surveys.:167 Some corrections that are needed for gravity anomalies are less important for magnetic anomalies. For example, the vertical gradient of the magnetic field is 0.03 nT/m or less, so an elevation correction is generally not needed.:167