MEASUREMENT 1

An overview of sensing platform-technological aspects for vectormagnetic measurement: a case study of the application in different

scenarios

Huan Liu1,2,3,4, Haobin Dong1,2,3, Jian Ge1,2,3, and Zheng Liu4

1School of Automation, China University of Geosciences, Wuhan, 430074, China2Hubei Key Laboratory of Advanced Control and Intelligent Automation for Complex Systems, Wuhan, 430074, China

3Engineering Research Center of Intelligent Technology for Geo-Exploration, Ministry of Education, Wuhan, 430074, China4School of Engineering, University of British Columbia Okanagan, Kelowna BC, V1V 1V7, Canada

Magnetic sensing platform techniques have been used in many years in an attempt to better evaluate the likelihood of recoverablehydrocarbon reservoirs by determining the depth and pattern of sedimentary rock formations containing magnetic minerals, suchas magnetite. Utilizing airplanes, large-area magnetic surveys have been conducted to estimate, for example, the depth of igneousrock and the thickness of sedimentary rock formations. In this case, the vector magnetic survey method can simultaneously obtainthe modulus and direction information of the Earth’s magnetic field, which can effectively reduce the multiplicity on data inversion,contribute to the quantitative interpretation of the magnetic body and obtain more precise information and characteristics ofmagnetic field resource, so as to improve the detection resolution and positioning accuracy of the underground target body. Thispaper presents a state-of-the-art review of the application situations, the technical features, and the development of the vectormagnetic sensing platform-technical aspects for different application scenarios, i.e., ground, wells, marine, airborne, and satellites,respectively. The potential of multi-survey sensing platform technique fusion for magnetic field detection is also discussed.

Index Terms—Magnetic survey technique, geophysical exploration, instrumentation, magnetometer, geomagnetic

I. INTRODUCTION

Magnetic sensing platform techniques are one of the mosteffective methods for geophysical engineering and environ-mental exploration [1]–[4], e.g., unexploded ordnance (UXO)detection [5], [6], mineral exploration [7]–[9], etc [10]–[13].In recent years, the vector magnetic survey techniques havedeveloped rapidly. When compared it with the traditionalscalar magnetic survey methods [14]–[16], the vector magneticsurvey can directly provide vector magnetic field informationand reveal detailed features of the Earth’s magnetic fieldwithout computing from the scalar measurements, which caneffectively reduce the multiplicity on data inversion [17]–[19].

For now, the Overhauser magnetometer and opticallypumped magnetometer are the two main scalar instruments.Without the deficiencies of a dead zone and heading errors thataffect optically pumped sensors, the Overhauser sensor haswider application prospects. Although scalar measurements arecharacterized by excellent performance, e.g., the resolution andthe accuracy of a commercial Overhauser magnetometer are0.01 nT and 0.1 nT, respectively, the magnetic informationobtained is limited. The vector magnetometers commonlyused in geophysical engineering mainly include fluxgate mag-netometers and superconducting magnetometers [20], [21].The fluxgate magnetometer is one of the most used vectorinstruments, with a resolution around 1 nT, which is worsethan that of the scalar instruments.

Since the vector magnetic survey data includes the magneticfield value and the corresponding direction, the data obtainedfrom the magnetic survey always needs to be corrected to

Corresponding author: Haobin Dong (email: donghb@cug.edu.cn).

the geographic coordinate system, corrected the drift withtemperature, etc., for further processing and interpretation. Forinstance, in the airborne magnetic survey, it is necessary toobtain the system attitude and orientation during flight, i.e., anerror of 10−3◦ of one of the inertial measurement unit anglecould produce an error of 1 nT for components after rotationform magnetometer coordinate system to the Earth coordinatesystem. Hence, the final magnetic measurement accuracyalways depends on the accuracies of the magnetometer andthe attitude measurement system, respectively [22], [23].

In literature [24], the total field anomaly can be transformedinto some other components of the magnetic field. However,such transformations will be limited for certain ambient-fielddirections in the same way that reduction to the pole is limitedas low latitudes. Consider, for example, a total-field anomalymeasured near the magnetic equator and caused by a bodywith purely induced magnetization. Both unit vectors in thedirection of the magnetization and in the direction of theambient field will have shallow inclinations in this case, andtransforming the total-field anomaly into the vertical compo-nent of the magnetic field can be expected to be an unstableoperation. Further, any noise within the measurements willgenerate artifacts, typically short in wavelength and elongatedparallel to the declination. Hence, the vector magnetic surveywould be a better choice in this case because it can obtain thevector magnetic field data directly.

To further suppress the uncontrollable influence, such asthe background noise, the geomagnetic diurnal change, etc.,the gradiometer based magnetic vector survey technique canbe employed to measure the difference between readingsfrom dual magnetometers. Using gradiometers we can simply

arX

iv:2

007.

0519

8v2

[ph

ysic

s.ge

o-ph

] 1

8 O

ct 2

021

MEASUREMENT 2

cancel out the spatially homogeneous part of the backgroundfield vector. Furthermore, the magnetic gradient is much lesssusceptible to orientation error compared to the magneticvector. To sum up, the vector magnetic survey, especially thecorresponding gradiometer based approach, can contribute tothe quantitative interpretation of the magnetic body and obtainmore precise information and characteristics of magnetic fieldresource, and thus improve the detection resolution and posi-tioning accuracy of the underground target body [25]–[27].

The vector magnetic survey technique can be mainly dividedinto ground magnetic survey, wells magnetic survey, marinemagnetic survey, airborne magnetic survey, and satellites mag-netic survey, and each technique has its own characteristics.The ground magnetic survey is mainly used for the detectionof ore bodies distributed horizontally on the surface [28].However, due to the abnormal superposition of different ge-ological bodies on the surface, it is difficult to distinguishthe depth. Hence, it is often combined with other magneticsurvey techniques [29]. The wells magnetic survey is basedon the magnetic characteristics of rock ore, measuring threeorthogonal components of the geomagnetic field, and the radialdetection range is large [30]. Further, this technique can beemployed to find both strong magnetite deposits and non-ferrous metals with weaker magnetic intensities, and it is aneffective method for detecting magnetic ore bodies, especially,provides a scientific basis for the exploration and evaluation ofmineral resources in the deep earth [31]. The marine magneticsurvey always adopts a ship equipped with magnetometersto conduct geomagnetic surveys in the ocean area. It playsa crucial role in the detection of military targets such asunderwater submarines, unexploded weapons, and magneticobstacles [32]. The traditional approach is mainly based ontotal field measurement.

In recent years, the three-component or full-tensor mag-netic gradient based marine magnetic survey technique hasbeen employed to obtain more geomagnetic information toprovide important parameters for the naval battlefield [33].The airborne magnetic survey is appropriate for scanning largeareas during reconnaissance to delimit target areas for detailedground surveys during the prospecting stage [34]. In particular,it can be used in coalfield studies to map out a broad structuralframework over an exploration area with complex terrain bothquickly and effectively [35]. The satellite magnetic surveycan obtain high-quality, global coverage magnetic survey data,conduct all-weather and uninterrupted measurements, and thuscarry out a series of scientific researches such as the evolu-tion of the geomagnetic field and the law of space currentmovement [36], [37].

In this paper, according to different application scenarios,first, the technical characteristics of the vector magnetic mea-surement compared with the scalar magnetic measurement aresummarized. Second, we elaborate on the application situa-tions, the technical features, and the instruments developmentof the ground vector magnetic survey technique, the wellsvector magnetic survey technique, the marine vector magneticsurvey technique, the airborne vector magnetic survey tech-nique, and the satellites vector magnetic survey technique,respectively. Finally, the advantages and disadvantages of these

five vector magnetic survey techniques are summarized, andtheir application prospects and development directions arediscussed.

II. APPLICATION IN DIFFERENT SCENARIOS

A. Ground vector magnetic measurement

The magnetic field at any point in space is a vector quantity,which means there is a direction associated with the field aswell as a field strength. The direction of the arrow can bethought of as the direction of the magnetic field. The lengthof the arrow can be thought of as the strength of the field, i.e.the longer the arrow, the stronger the field. We can define thislength as B, which implies that B represents the strength ofthe magnetic field. Generally, a single axis measuring devicewill change its reading depending on which way the sensitiveaxis is oriented with respect to the direction of the magneticfield. To obtain a complete representation of magnetic field atany point in space, one needs not only the value of B, but thedirection, which can be expressed as the three components,Bx, By and Bz . Some magnetic field sensors measure onlyone component of the magnetic field, e.g., Fluxgates andHall effect instruments which are referred to as single axisdevices [38], [39]. Other instruments measure only the totalfield B, e.g., NMR [40], [41], ESR [42], [43]. Aiming to getmore information for engineering geophysical exploration, itis possible to combine single-axis sensors to give three fieldmeasurements in a single probe package. These are referredto as three-axis devices [44], [45].

The ground vector magnetic survey using three-axis deviceis mainly divided into two types of measurement: 1) stationarymeasurement and 2) mobile measurement. The purpose isto analyze the geomagnetic characteristics by measuring theEarth’s magnetic field. The main task is to survey ore bodieswith magnetic anomalies, geological structural zoning, andgeological mapping services [46]–[50].

The stationary measurement has been widely used inmany fields such as the basic network of seismic stations,underground magnetic anomaly source detection, and weakmagnetic resonance signal acquisition, because of its earlydevelopment and easy implementation [51], [52]. The basicstation network of the China Earthquake Administration’sgeomagnetic station has been using fluxgate magnetometersto detect changes in the geomagnetic field in seismic dangerzones since 2003 [53], [54]. This method can be used todetect earthquake anomaly precursors in time and providebasic information for accurate prediction of earthquake oc-currence. Meanwhile, the observed geomagnetic field strength,inclination, and declination can be adopted to study the Earth’sbasic magnetic field and deep underground structures [55].

In addition, the high-quality observation data of the iono-sphere, the substorm current, and the very-low-frequency(VLF) magnetic field of the measured region can be used tostudy the space weather [40]. For the particularity of observa-tions, the fluxgate magnetometer array system can realize theNetwork functions such as communication access, automaticidentification, monitoring & management, and data collectionfor field mobile observation equipment. According to the

MEASUREMENT 3

temporal and spatial distribution characteristics of the geo-magnetic field, each three-component fluxgate magnetometerwhich is arranged in the basic network channels, is connectedto the monitoring center through the wireless or wired Internetnetwork, and a large amount of geomagnetic raw observationdata is acquired after a period of recording to achieve reliableand real-time monitoring of the geomagnetic data. This systemplays an important role in the observation platform for short-period geomagnetic. However, the observation interruptionwill occur due to some factors such as the observation site,power supply, communication conditions, etc., and its stabilityneeds to be further strengthened. Hence, a SuperconductingQuantum Interference Device (SQUID) could be considered tomeasure the variation of the three-component magnetic fieldto obtain a higher sensitivity [56], [57].

Fig. 1: JESSY SMART ground tensor magnetic gradientmeasurement system developed by IPHT [58].

For the mobile measurement, the Institute for PhotonicTechnologies (IPHT) of Jena, Germany, adopted a low tem-perature superconducting (LTS) SQUID to develop a groundtensor magnetic gradient measurement system, named JESSYSMART [58], as shown in Fig. 1. The JESSY SMART isa novel ground tensor magnetic gradient measurement sys-tem for fast three-dimensional (3D) geomagnetic mapping ofhidden subsurface anomalies. It can localize smallest Earth’smagnetic field gradient inhomogeneities in a depths of up toapproximate 10 meters. The sensors are mounted onto a cartwhich can be moved attached to a motor car or alternativelypushed along manually. It has the following advantages whencompared with conventional geomagnetic methods: 1) highestpossible magnetic field resolution (sensor 7 fT/cm); 2) fastmapping of measuring area (up to 7 acre and accordingly 3ha/hour); 3) very accurate position and spatial resolution ofdata in cm range); 4) measurement in difficult ground, envi-ronment, and climate conditions possible; 5) nondestructivemeasuring and a high planning reliability.

Figure 2 shows the measurement results from a 400-meterlength area in an urban street of the city center, Plauen,Germany. The mission is to investigate and clear the UXO

Fig. 2: Investigation and clearing of UXO before road con-struction using the JESSY SMART [59].

before road construction. To minimize the road traffic andconstruction work disturbing influence, the mission was im-plemented at night time. The magnetogram shows a coupleof extended anomalies which correspond to various metallicobjects assembled in refilled pits, a typical residual of the veryfirst simple clearance procedures carried out directly before theconstruction. The magnetic anomalies and the correspondinglocalization of the magnetic objects can be visualized clearly,which demonstrates the effectiveness of the system.

In 2017, the Strasbourg Institute of Globe Physics [60]developed a fluxgate based mobile measurement system forgeoreferenced magnetic measurements at different scales, i.e.,from sub-metric measurements on the ground to aircraft-conducted acquisition through the wide range offered by un-manned aerial vehicles (UAVs), with a precision in the order of1 nT. Such equipment can be used for different kinds of appli-cations, such as structural geology, pipes, and UXO detection,archaeology. Likewise, Jilin University proposed a novel noisereduction method and developed a corresponding magneticresonance three-component noise reduction device [61], [62].However, this method is in the laboratory stage, and itspracticability for field application still needs to be furthervalidated.

B. Wells vector magnetic measurement

Wells vector magnetic survey technique is used for cen-sus exploration of magnetite deposits or poly-metallic de-posits containing ferromagnetic minerals. It is a geophysicalprospecting method for the development and extension ofthe ground magnetic surveys in space, based on the studyof the magnetism of rocks and ore bodies. Besides, thistechnique has a unique advantage in the exploration of deepmagnetic deposits, and it can solve some problems that thesurface magnetic survey cannot solve, for instance, the depthposition of the ore tail and the top of the mine. The wellsmagnetic survey can avoid the human-made interferences andthe heterogeneity of shallow magnetic sources, and the verticalresolution is relative high. Because of these advantages, thethree-component magnetic survey in the well has become aneffective method for surveying every hole of the magnetitedeposit in the general survey [63].

The three-component magnetic survey in the well is carriedout along the drilling direction of the borehole. Throughmeasuring the x, y, and z components of the magnetic field atdifferent depths in the well, the magnetic anomaly components∆x, ∆y, and ∆z can be calculated to judge and evaluate the

MEASUREMENT 4

blind magnetic ore beside the well, the blind ore at the bottomof the well, the shape of the ore body, the size of the mineral,etc [64]–[66]. The three components ∆x, ∆y, and ∆z canalso be synthesized into an total intensity anomaly ∆T .

In practical applications, aiming to facilitate the qualitativestudy of magnetic anomalies, ∆T is divided into a verticalcomponent ∆Z and a radial component ∆H , and the ∆Hcan be further divided into a horizontal component which isperpendicular to the direction of the ore body and a horizontalcomponent which is parallel to the direction of the ore body.In this case, we can obtain the relations between the magneticanomaly strength and the direction of the ore body. Throughthe analysis of the parameters such as the magnetic verticalcomponent and the radial component of the magnetic anomalyarea, we can infer whether there are ore bodies at the bottomof the well or besides the well.

Fig. 3: Overall view of the KTB drilling system and thecorresponding module structure [67].

In the mid-1960s, Wang et al. [68] developed China’sfirst three-component magnetic survey system for transistorwells. This system can directly measure the three-componentmagnetic field value of the vertical coordinate system in thewell, using a gravity orientation with two degrees of freedom.When compared with the degree of freedom (DOF) gravityorientation system developed in Sweden at that time, it hasobvious advantages such as high measurement accuracy, real-time measurement, etc.

In the 1990s, the German Continental Drilling Program(KTB) has made great progress in data processing and in-version interpretation of magnetic surveys in boreholes [69].Fig. 3 shows the overall KTB drilling system [67], in whicha combination of a closely spaced surface gravity survey witha high-resolution helicopter aeromagnetic survey as well asborehole gravity and magnetometer measurements allowed adetailed 3D modeling of the anomalies at the KTB drill site.

To be specific, in the KTB main hole, numerous fluxgate mag-netometers are equipped in the vertical drilling system (VDS),collecting the magnetic field data. The magnetic measurementsare always carried out at depths from 300 m to 6000 m, andthe errors are less than 10 nT in total after data processingincluding special correction, sophisticated calibration, etc.For a better resolution of the magnetic anomaly, a detailedhelicopter survey is carried out with the drill site in its center.Radiometric and electromagnetic data are also collected. Inthis case, an almost complete coverage and high resolution ofthe magnetic field in the measuring level is obtained, yieldingan excellent data set for analyzing the geological structure andsupporting surface geological mapping.

A surprising result using the KTB is an anomalous verticalgradient of the geomagnetic field which was detected in thelower part of the borehole down to at least 6000 m. Toemphasize the general trend of the magnetic field in theborehole, the data were processed with a running 40 m low-pass median filter. The results are shown in Fig. 4. Withoutany magnetized bodies at depth, a vertical gradient of the totalmagnetic field of about 22 nT/km should be expected from thelocal international geomagnetic reference field (IGRF). Thesource bodies in the upper part of the borehole produce evena decreasing field in the lower part of the drill hole. However,gradients of up to 200 nT/km have been measured. The averagegradient below 3000 m is 60 nT/km. Below 6000 m the trendsand gradients of the magnetic field could not be determinedbecause of the low accuracy of the orientation tool.

Fig. 4: The trend of the horizontal and vertical component ofthe magnetic field in the KTB main drill hole [70].

Moreover, the United States, Sweden, and Canada havemade important contributions to the design of magnetic mea-suring instruments, data processing, etc. For instance, Kuhnke

MEASUREMENT 5

et al. [71] designed a borehole magnetometer, analyzed thepossible problems of the magnetometer during operation inthe well, and developed different types of instrument systemsbased on the temperature resistance characteristics. Bosum etal. [70] analyzed the characteristics of borehole magnetic sur-vey data, and Worm [72] conducted rock magnetic simulationstudies by combining with KTB borehole magnetic surveydata. Leonardi et al. [73] adopted KTB magnetic survey datato study the anisotropy of the crustal structure, and obtainedsome fractal features.

In the early 21st century, a new three-component mag-netometer, dubbed Goettinger borehole magnetometer whichhas been successfully used to measure in the OutokumpuDeep Drill Hole (OKU R2500, Finland), the repeatability ofthe magnetic field vector is 0.8◦ in the azimuthal direction,0.08◦ in inclination and 71 nT in magnitude [74]. More-over, a reoriented vector borehole measurement was carriedout using the Goettinger magnetometer at two sites duringIntegrated Ocean Drilling Program (IODP) Expedition 330to the Louisville Seamount Chain [75]. Likewise, ChongqingGeological Instrument Factory developed a high-precisionthree-component magnetometer, named GJCX-1 [76]. Theorientation differences of x and y direction are both lessthan 100 nT, the orientation difference of z direction is lessthan 50 nT. Shanghai Institute of Geosciences developed ahigh-precision three-component logging tool, named JCC3-2A, in which the three-axis giant magnetoresistive sensor isused as the magnetic sensitive element, and the three-axisgravity acceleration sensor is used as the directional element.The sensitivity of the magnetic component is 40 nT, and theaccuracy is better than 100 nT [77].

Consequently, the wells three-component magnetic surveytechnique has the advantage of a large radial detection range,which provides an effective tool for dividing the magnetic rockor ore body on the borehole profile, the magnetic mineralsbesides or below the exploration well, and the blind ore bodiesassociated with the magnetic minerals. To a certain extent,it is possible to perform depth calibration on the inversionof the ground magnetic survey, solving the detection workthat cannot be completed by the ground magnetic surveytechnique, which implies that the wells vector magnetic surveyan effective method for deep ore body detection. However,due to the limitations of borehole diameter, well temperature,and the size & volume of the instrument itself, the accuracyof the magnetic sensor used in the wells magnetic surveyis lower than that of the ground magnetic survey, and theoverall observation accuracy is also lower than the surfacemagnetic survey by an order of magnitude. Hence, it is of greatsignificance to develop a higher-precision three-componentmagnetic measuring instrument in the well.

C. Marine vector magnetic measurement

Marine magnetic survey is the main approach to obtainthe information on the distribution and variation character-istics of the geomagnetic field in the ocean area, and itis also one of the important cases of marine engineeringsurveys and military hydrographic surveys [78]. In the 1950s,

Vacquier et al. [79] adopted a fluxgate magnetometer tomeasure the geomagnetic field in the three oceans. Sincethe 1980s, the international geoscience community has begunto implement seafloor geomagnetic observations, deployingdozens of seafloor observation stations to conduct oceanicand geomagnetic surveys, and carrying out research on un-derground deposits and geodynamics. Afterwards, the earth-based electromagnetic measurement method based on artificialexcitation appeared, e.g., electromagnetic method of oceancontrolled source [80].

Up to now, the marine magnetic survey mainly employs anoptical pump magnetometer or an Overhauser magnetometerfor the underwater magnetic survey [81]–[83]. To suppress theinfluence of waves, the magnetic probes are usually measuredat a certain depth below the surface of the ocean, using a drag-and-drop measurement approach. In 1997, Seama et al. [84]developed a vector magnetometer for deep tow detection. Todetermine the position of the tow body in the deep ocean, aninertial navigation system and GPS, short baseline acousticmeasurement and pressure measurement were adopted. Geeand Cande [85] developed a vector magnetometer system witha speed of 10 ∼ 12 knots. The test results show that the devicecan determine the horizontal and vertical components with anaccuracy of 30 nT ∼ 50 nT, which is of great significance formagnetic survey applications in low latitude areas.

To further identify the magnetic anomalies caused by short-polarity events or other local magnetic bodies, Blakely etal. [86] developed a magnetic vector measurement system.Besides, Horner and Gordon [87] adopted spectral analysismethods to identify anomalies in the seabed magnetic bands,and they found that the aeromagnetic survey data had betterresults than shipboard total field survey data.

In 1995, the Woods Hole Oceanographic Institute developedthe world’s first unmanned submersible autonomous benthicexplorer (ABE) for marine magnetic survey [88], which ismainly composed of conductivity probe, temperature probe,depth gauge, camera and magnetometer, as shown in Fig. 5.The ABE is a robotic underwater vehicle used for exploringthe ocean to depths of 4,500 meters. It was the first AUVused by the U.S. scientific community. ABE is often used intandem with Alvin or Jason surveying large swaths of oceanfloor to determine the best spots for close-up exploration. ABEis designed to perform a pre-programmed set of maneuvers,using its five thrusters to move in any direction, hover, andreverse. The AUV excels at surveys of the shape of theseafloor, its chemical emissions, and its magnetic properties.ABE is particularly valuable in rugged terrain. On-boardsensors tell the vehicle how deep it is and how far it is offthe ocean floor, and the AUV calculates its horizontal positionby contacting a system of acoustic beacons (transponders) setout in fixed locations [89]. In 2014, the 715-research institutedeveloped the RS-HC3 ocean tensor magnetic gradiometerwhich dynamic range is -100000 nT ∼ 100000 nT, and thissystem also has got a preferable result.

Figure 6 shows the fault inference in the Longqi hydrother-mal area based on near-bottom magnetic data collected by theABE 200 dive. The blue circles point location of the moundsand the solid red lines indicate the trap areas. There is an

MEASUREMENT 6

Fig. 5: Autonomous underwater vehicle ABE [88]

obvious high magnetic anomaly on the top of the main moundand only a weak positive anomaly on the secondary mound,because the anomaly resolution decreases with increasingdepth. In addition, the corresponding ISDV values are well-defined for the main and secondary mounds with the valuesshowing a well-defined trap area and a smaller secondary trap.

Fig. 6: Fault inference in the Longqi hydrothermal area basedon near-bottom magnetic data collected by the ABE [90]. (a)Near-bottom magnetic field; (b) The spatial differential vector(ISDV) distribution.

D. Airborne vector magnetic measurement

The resource conditions and environmental conditions inmany areas of the earth are complex, such as swamps, forests,deserts, and mountain areas, where are inconvenient for per-sonnel equipment to access [91], [92]. The airborne electro-magnetic detection system can overcome the impact of thecomplex ground environment. It has the characteristics of lowcost, fast survey speed, wide survey area, and strong versatility,and plays an important role in geophysical exploration [93],[94]. Before the 21st century, the airborne magnetic surveytechnique is mainly based on the measurement of total fieldstrength or gradient. As the continuous development of mag-netic exploration theories and methods, the airborne surveytechnique has been promoted from the original total fieldstrength or gradient measurement to vector (three-componentor gradient tensor) measurement [95], [96].

In 2003, the Australia BHP Billiton Petroleum Companydeveloped a fluxgate aviation three-component magnetic mea-surement system, and conducted flight tests in the Rocky

Strip iron-bearing construction area of Western Australia [97].After attitude correction, three-component geomagnetic fielddata and attitude were obtained, in which the corrected noiselevel is 50 nT ∼ 100 nT. In 2011, Munschy et al. [98]installed two fluxgate magnetometers on the tail of the MauleMX7 small aircraft, and conducted flight tests in the Vosgesarea. After magnetic compensation, the magnetic field dataof two horizontal components were obtained, and the verticalcomponent was calculated by combining the measurementresults of the total field. After mapping the magnetic fielddistribution in this area, it is basically consistent with theactual geological conditions. In the same year, the ChinaAero Geophysical Survey and Remote Sensing Center (AGRS)developed an airborne three-component magnetic survey sys-tem, which is composed of a fluxgate magnetometer andan INS/GPS strap-down inertial navigation module [99]. Inaddition, a verification test was conducted by mounting thissystem to a fixed-wing aircraft [100].

At the beginning of the 21st century, another airborne vectormagnetic survey technique: aeronautical full-tensor magneticgradient measurement has become a hot spot for geophysicistsin various countries. This technique refers to measure thespatial change rate of the three field components of thegeomagnetic field vector along three mutually orthogonalaxes, and there are nine elements in total [2]. Its outstandingadvantages include the invariant contour plot calculated fromthe gradient tensor is easy to explain, the dipole trackingalgorithm can be used to accurately determine the depthand horizontal position of the magnetic dipole, etc. Hence,the high-precision 3D positioning of underground magneticgeological bodies and ore bodies can be realized through full-tensor magnetic gradient measurement data, and their spatialdistribution information can be obtained [4].

In 2003, the U.S. Geological Survey (USGS) conductedproof-of-principle tests of the tensor magnetic gradiometersystem (TMGS) at the Strategic Environmental Research andDevelopment Program’s (SERDP) UXO Standardized TestSite at the Yuma Proving Ground (YPG), Arizona [101]. Thenoise of this system in 0.001 Hz ∼ 100 Hz band is typicallyless than 0.1 nT, and the thermal stability is about 0.2 nTdeg−1 for a field of 60000 nT [102]. The objective of thesetests was to assess the potential of this system for UXOapplications and to help us formulate design parameters foran improved TMGS designed specifically for UXO. Althoughthe current system was not designed with UXO applicationsin mind, it detected the majority of the buried targets in theCalibration Area at YPG. Its performance rivaled that of thecesium vapor companion system, and it showed true tensorgradiometric performance over a selected target.

Figure 7 shows a tensor invariant map over a 60 mm mortarround buried 0.25 m deep of the YPG measuring by theTMGS. The array was towed over the grid at an average heightof 0.3 m, and most of the measured magnetic gradients greatlyexceed 2 nT/m. The tensor invariant maps have been despikedand corrected for GPS lag, and system timing latency errors.The tensor anomaly is positive and highly peaked over thetarget at the center of the grid. The positive anomalies at theedges of the map correspond to adjacent off-grid targets.

MEASUREMENT 7

Fig. 7: The tensor invariant map over a 60 mm mortar roundburied 0.25 m deep of the YPG measuring by the TMGS [101].

Jilin University constructed a magnetic full-tensor gra-diometer, which utilizes four fluxgates arranged on a planarcross structure, and a single, triaxial, spherical feedback coilassembly [103]. In this arrangement, one of the fluxgatesis used as a reference, controlling the currents through thefeedback coils. Since the fluxgates are working in the near-zeromagnetic field environment, the magnetic tensor gradiometeris stable and of an improved accuracy. However, due to the lowsensitivity of the fluxgate magnetometer, the baseline distancebetween the sensors is very large when performing gradientmeasurement, and thus, the requirements on the system struc-ture are relative high, and the measurement accuracy is difficultto guarantee. Consider for this case, the SQUID based vectormagnetic sensor with higher sensitivity has become the firstchoice for the magnetic gradient tensor measurement system.

In 2005, the U.S. Oak Ridge National Laboratory (ORNL)developed an aeronautical full-tensor magnetic gradient mea-surement system which is composed of high temperature su-perconducting (HTS) SQUIDs and an aeronautical geophysicalplatform [105], and implemented a ground experiment for thissystem. The system adopts eight HTS-SQUID sensors whichare all fixed on a bracket, forming a full-tensor probe througha reasonable layout.

Likewise, the Commonwealth Scientific and IndustrialResearch Organization (CSIRO) developed an HTS-SQUIDbased areo full-tensor magnetic gradient measurement system,namely, UXOMAG, as shown in Fig. 8. The UXOMAGis a mobile magnetic tensor gradiometer prototype systemcapable of not only detecting, but also locating and classifyingUXO. This system can measure all five independent magnetictensor components allowing the user to discriminate betweenharmless metallic debris or shrapnel and a potentially lethalUXO [104]. It uses multiple CSIRO’s HTS planar SQUID

gradiometers configured unique geometry to determine themagnitude of the full-tensor. Referencing magnetometers areused to improve common mode rejection. The UXOMAG’shigh sensitivity has the potential to detect 40 mm calibre UXOat a distance of up to 4 m away.

There are also some other SQUID based full-tensor mag-netic gradient measurement systems, e.g., the prototype devel-oped by Jilin University [107]. For this system, six first-orderplanar-type SQUID gradiometers mounted on the six slopingfacets of a hexagonal pyramid, which can form the probeand determine all independent components of the magneticgradient tensor with the help of triaxial SQUID magnetometersplaced near the probe. The gradiometer sensitivity is about 100pT/m. The Shanghai Institute of Microsystem and InformationTechnology (SIMIT) and the Institute of Remote Sensing andDigital Earth (RADI) from the Chinese Academy of Sciencesjointly developed a prototype of HTS-SQUID based areo full-tensor magnetic gradient measurement system, and conducteda flight measurement test in Inner Mongolia of China. Thegradiometer sensitivity is up to 50 pT/m. However, these twosystems are still in the experimental stage.

Up to now, the most commonly used and accepted practicalaero full tensor magnetic gradient system is developed byIPHT. The IPHT won the ”Outstanding Achievement Award”in the category Mining Research for their ground breakingresearch in the field of world’s first full-tensor airborne mag-netic gradiometer, known by its synonym ”JESSY STAR”.Fig. 9 illustrates the overall system. The JESSY STAR systemhas surpassed all conventional surveying technology in perfor-mance and, in addition, provided the exploration communitywith magnetic data never measured before. It also utilizes theultimate sensitivity of LTS-SQUID sensors [108]. The JESSYSTAR system include six gradiometers and three magnetome-ter channels, to compensate for sensor motion noise. Further,high spatial resolution is provided by a differential GPS systemusing SBAS and an inertial unit (INU), which are synchronizedwith the magnetic gradient data, The INU data is used for mo-tion compensation. The data acquisition system, incorporatea small-sized 15 channels 24 bit analogue digital converterunit. To date, JESSY STAR has flown hundreds of hours inboth helicopter mode and fixed wing plane configuration, andit has transformed exploration from ”flying blind” to ”seeingclearly”. This new generation system is capable of detectingminerals and precious metals that were deemed ”undetectable”before.

The magnetic measurements with a UAV are ideal for fillingthe gap between the ground and airborne magnetic surveying.For several years, there has been a very important developmentof magnetic measurement with UAVs by using the scalar orvector magnetometers. However, for now, although the vectormagnetic measurement has numerous advantages, the tech-nological limitations in sensor manufacturing and unwantedmagnetic fields will corrupt the measurements of three-axismagnetometers, especially in the airborne or satellite vectormagnetic survey. Since the three-axis magnetometer (take thefluxgate as an example) is not an absolute instrument, it hasto be calibrated and compensated.

Generally, there are two different principles for calibrating

MEASUREMENT 8

Fig. 8: HTS full-tensor magnetic gradiometer developed by CSIRO [104]. (a) The magnetic probe’s structure; (b) The magneticprobe; (c) The overall system; (d) System experiment on ground.

a magnetometer. In a vector calibration, the output of thevector magnetometer is compared with the known magneticfield vector which is applied to the instrument. The prob-lem of an in-flight vector calibration is of course that theapplied magnetic field vector is in principle unknown. In ascalar calibration, only the intensity of the magnetic field isused, but not its direction. However, the scalar intensity isknown from the second magnetometer, e.g., the Overhauserscalar magnetometer. Calibration of the vector magnetometeris based on the following points: 1) use pre-flight values forthe temperature dependence; 2) use scalar in-flight calibrationfor the estimation of the nine intrinsic parameters of thevector magnetometer; 3) use vector in-flight calibration forthe simultaneous estimation of the remaining three calibrationparameters together with a spherical harmonic model of theEarth’s magnetic field. More details can be refereed to litera-ture [109], [110].

Likewise, Gavazzi et al. [111] implemented a case study onthe aerial military base BA112 shows the usefulness of theinstruments for the detection of underground pipes, UXO, andarchaeological remains. Maire et al. [112] proposed a newmethod for rapid mapping and upscaling from the field toa regional scale using a UAV and a fluxgate magnetometer.However, to obtain accurate aeromagnetic data, the compen-sation of the magnetic effects of the UAV is a challenge. Most

research on aeromagnetic compensation has mainly focusedon the suppression of the magnetic interference generated byaircraft maneuvers and the on-board electronic equipment.Amount of compensation methods, such as ridge regres-sion [113], truncated singular value decomposition [114], re-cursive least-squares [115], partial least-squares [116], supportvector methods [117], c-k class estimation [118], and princi-pal component analysis [119] have been proposed one afteranother, ensuring the high-quality of measured aeromagneticdata.

The inertial measurement units (IMU) is an affordableinstrument for the determination of orientation. The fluxgatesensors embedded in the system are affected by nonidealitiesthat can be greatly compensated by proper calibration, bydetermining sensor parameters, such as bias, misalignment,and sensitivity/gain. For instance, an online calibration methodfor a three-axial magnetometer using a 3D Helmholtz coil isproposed in the literature [120]. The magnetometer is exposedto different directions of the magnetic field created by the3D coil. The parameters are estimated by using an unscentedKalman filter. The directions are calculated online by using asensor parameter covariance matrix.

In addition, ferromagnetic interferential field from platformsis one of the most dominating error sources for magnetometer.For magnetic vector and gradient tensor measurement, what

MEASUREMENT 9

Fig. 9: Setup of the airborne SQUID system developed by IPHT [106].

is cared about most is the effect of compensating magneticfield vector. For instance, a novel compensation method hasbeen proposed in literature [121], in which attitude informationis used to set up a vectorial compensation model and leastsquare method is used to estimate parameters. Fig. 10 showsthe uncompensated and compensated results. It can be seenthat the measurement errors of geomagnetic field vectors andmagnitude decrease remarkably after compensation.

Fig. 10: Uncompensated and compensated results [121].

E. Satellites vector magnetic measurement

In 1958, the Soviet Union launched the first satelliteSPUTNIK-3 to measure the geomagnetic field, which fore-shadowed the beginning of the satellites magnetic survey

technique [122], [123]. The fluxgate magnetometer was in-stalled on this satellite, however, since the direction of themagnetometer cannot be accurately determined, only totalfield strength data were obtained. Since then, a series ofmagnetic survey satellites carried total field magnetometerswere launched by the Soviet Union, including proton pre-cession magnetometers or optical pump magnetometers. Like-wise, they cannot be called vector magnetic survey. With thedevelopment of space magnetic field measurement technology,there have been numerous professional satellite programsdedicated to the mapping of the Earth’s intrinsic geomagnetic,such as the U.S. MAGSAT satellite program [124], [125],Denmark’s Ørsted satellite program [126], [127], Germany’sCHAMP satellite program [128], [129], Argentina’s SAC-Csatellite program [130], and European Space Agency’s (ESA’s)Swarm [131].

The MAGSAT spacecraft was launched into a twilight, sun-synchronous, orbit with inclination 96.76o, perigee 352 kmand apogee 561 km [132]. The Cesium vapor scalar andfluxgate vector magnetometers together measured the fieldmagnitude to better than 2 nT and each component to betterthan 6 nT [133]. Two star cameras, a high-accuracy sunsensor and a pitch axis gyro provided the 10-20 arc-secondattitude measurements necessary to achieve this accuracy. Themagnetometers were located at the end of a boom to eliminatethe effect of spacecraft fields. An optical system measured theattitude of the vector magnetometer and sun sensor (at theend of the boom) relative to the star cameras (on the mainspacecraft) [134]. The data are available in several formatsfrom the National Space Science Data Center and are under-going analysis by a team of investigators [135]. In addition,An et al. [136], [137] employed the method of spherical capharmonic analysis to derive a spherical cap model based onthe MAGSAT data set, and described the three-dimensional

MEASUREMENT 10

structure over Asia and Europe region, respectively.

Fig. 11: Illustration of the Ørsted spacecraft in orbit [138].

The Ørsted satellite, named after the Danish scientist HansChristian Ørsted, is the first satellite mission since MAGSATdesigned for high-precision mapping of the Earth’s magneticfield [109]. It was launched with a Delta-II rocket fromVandenberg Air Force Base into a near-polar orbit. As thefirst satellite of the ”International Decade of Geopotential Re-search”, the satellite and its instrumentation have been a modelfor other present and forthcoming missions like CHAMP andSwarm [138]. The Ørsted spacecraft is composed of two maininstruments, including an Overhauser magnetometer and acompact spherical coil based triaxial fluxgate magnetometer,as illustrated in Fig. 11. The objective of the Overhausermagnetometer is to measure magnetic field scalar values withan absolute measurement error better than 0.5 nT, an dynamicrange from 16000 nT to 64000 nT, and a sampling rateof 1 Hz [139]. The fluxgate magnetometer which measuresmagnetic field vectors at an absolute measurement error lessthan 1 nT, an dynamic range from -65536 nT to 65536 nT, anda resolution better than 0.25 nT. Further, the collected vectordata is calibrated using the absolute intensity measured bythe Overhauser magnetometer. After calibration, the agreementbetween the two magnetometers is better than 0.33 nT [140].

Figure 12 illustrates the geomagnetic field of two periods(1980 and 2000) recorded by the Ørsted satellite. The topfigure displays a color-coded global image of the magneticfield strength in the year 2000 as modelled from the Ørsteddata. The scale spans from 20000 nT ∼ 70000 nT. The bottom

Fig. 12: Illustration of the geomagnetic field of two periods(1980 and 2000) [141].

field displays the differences between Ørsted results from 2000and results from the MAGSAT mission. The scale spans herefrom - 2400 nT to + 1800 nT. The changes in field strengthover the 20 years span between the two missions are mostlynegative and ranges up to almost 10% of the total field.

Fig. 13: Photograph of the CHAMP satellite [142].

The CHAMP is a German small satellite mission for geosci-entific and atmospheric research and applications, managed byGeoForschungsZentrum (GFZ) [143], [144]. The satellite wasbuilt by the German space industry with the intent to fosterhigh-tech capabilities, especially in the East-German spaceindustry, and it was launched from the cosmodrome Plesetsk(north of Moscow) aboard a Russian COSMOS launch vehi-cle [145]. Fig. 13 shows a photograph of the CHAMP satel-lite. The payload includes an accelerometer, a magnetometerinstrument assembly system, a GPS receiver, a laser retroreflector, and an ion drift meter. The magnetometer instrumentassembly system is a boom-mounted package consisting of anOverhauser scalar magnetometer (the measurement range is16000 nT ∼ 64000 nT, the resolution is 0.1 nT, the absoluteaccuracy is 0.5 nT, and the sampling rate is 1 Hz.) [146],two fluxgate vector magnetometers (the measurement range

MEASUREMENT 11

is -64000 nT ∼ 64000 nT and the resolution is 1 nT ∼ 2nT) [147], and two-star imagers to provide attitude informationfor fluxgates [148].

Fig. 14: Artist’s view of the deployed SAC-C spacecraft [149].

The SAC-C satellite (illustrated in Fig. 14) is an interna-tional cooperative mission between the National Aeronauticsand Space Administration (NASA), the Argentine Commissionon Space Activities (CONAE), the French Space Agency, theBrazilian Space Agency, the Danish Space Research Institute,and the Italian Space Agency [150]–[152]. The SAC-C was de-veloped through the partnership of its senior partners, CONAEand NASA with contributions from Brazil, Denmark, France,and Italy. It was the third satellite launched by CONAE andwas the first operational Earth observation, designed to meetthe requirements of socio-productive areas of our country,including agriculture, hydrology, coastlines, geology, healthemergencies, etc. This satellite is equipped with a vectormagnetometer and a star imager, further included are theassociated support electronics to control the sensors and theboom deployment, a power control unit and a command anddata handling unit. The geomagnetic measurement system onSAC-C represents essentially the same compact spherical coilbased vector magnetometer as those flown on the Ørstedand CHAMP missions [153]–[155]. Besides, a scalar Heliummagnetometer is placed on the tip of the boom and associatedsupport electronics completes the magnetic mapping payload.The magnetic field measurements have a resolution of 1 nT atscalar and 2 nT at vector [156].

The Swarm satellite is constructed for the first constellationmission of the European Space Agency (ESA) for geomag-netic observation [158]. This mission is operated by ESA’sEuropean Space Operations Centre (ESOC), in Germany, viathe primary ground station in Kiruna, Sweden. The researchobjectives of the Swarm mission is to provide the best-eversurvey of the geomagnetic field. Hence, a satellite constellationstrategy is adopted. The satellite constellation is a group ofartificial satellites working together as a system. Unlike asingle satellite, a constellation can provide permanent globalor near-global coverage, such that at any time everywhere on

Earth at least one satellite is visible. Satellites are typicallyplaced in sets of complementary orbital planes and connect toglobally distributed ground stations. They may also use inter-satellite communication. This mission consists of the threeidentical Swarm satellites (A, B, and C), which were launchedinto a near-polar orbit [159]. All the three Swarm satellitesare equipped with the following set of identical instruments,including an absolute scalar magnetometer, a vector field mag-netometer, a star tracker, an electric field instrument, a GPSreceiver, a laser retro-reflector, and an accelerometer [160],[161], as shown in Fig. 15. The absolute scalar magnetometeris an optically-pumped meta-stable helium-4 magnetometer,developed and manufactured by CEA-Leti under contract withthe French Space Agency. It provides scalar measurements ofthe magnetic field to calibrate the vector field magnetometer.The vector field magnetometer is the mission’s core instru-ment, which is developed and manufactured at the TechnicalUniversity of Denmark [162]. It makes high-precision mea-surements of the magnitude and direction of the magneticfield. The orientation of the vector is determined by the star-tracker assembly, which provides attitude data. The vector fieldmagnetometer and the star-trackers are both housed on anultra-stable structure called an optical bench, halfway alongthe satellite’s boom.

III. DISCUSSION

As the basic method of the vector magnetic survey tech-nique, the ground vector magnetic survey development is theearliest and the technique is the most mature. The accuracy ofthe instrument system for the ground vector magnetic surveyis high and the working performance is reliable. Nevertheless,there are also some problems, e.g., the detection depth isshallow and working efficiency is low. The wells vectormagnetic survey is one useful tool for the Earth’s deep mineralresources exploration. It can finish the detection work that theground vector magnetic measurement can’t do. However, dueto the limit of the bore diameter and the high temperature inthe borehole, the detection accuracy of the instrument systemis lower than the ground magnetic instrument. The marinevector magnetic survey has an irreplaceable effect in theapplication of military or other ocean engineering. Its workingmode changes from the method of ship drag-and-drop to smallunmanned underwater vehicle, which can measure a moreprecise underwater magnetic anomaly. The airborne vectormagnetic survey has the superiority of fast detection, highefficiency and strong practicality to the complex geophysicalenvironment, etc. However, the detection accuracy is relativelylow because of magnetic interference, the posture change ofthe aircraft carrier, etc. The study of magnetic compensationand data processing methods for aero three-component andfull-tensor magnetic gradient is also needed. The satellitesvector magnetic survey can obtain high-quality and wholeglobal magnetic field data. It has the advantages of high de-tection efficiency, wide detection range, etc., accelerating andfacilitating the investigations on space measurement techniqueand the evolution of the Earth’s magnetic field.

In recent years, the multi-survey technique fusion ap-proaches for geomagnetic surveys get a lot of attention, such

MEASUREMENT 12

Fig. 15: Illustration of the Swarm satellite in orbit [157].

as the aforementioned KTB drilling system. In the borehole,the fluxgate magnetometer is employed to collect the magneticfield data. For a better resolution of the magnetic anomaly, adetailed helicopter survey is carried out with the drill site inits center. In this case, complete coverage and high resolutionof the magnetic field in the measuring level is obtained, whichcan obtain an excellent data set for analyzing the geologicalstructure and supporting surface geological mapping. Fur-ther, there are some multi-survey data fusion based inversionmethods, such as the joint inversion of surface and three-component borehole magnetic data [163], the joint inversion ofelectromagnetic and magnetic data [164], etc., which can fur-ther improve the detection resolution for near-surface studies.However, there are still some problems with the multi-surveytechnique fusion that need to be further considered. To bespecific, different magnetic vector survey techniques have towork simultaneously to achieve real-time synchronization ofdata measurement. In this case, these systems will be interferedby each other, causing the data abnormality. Furthermore,the problems of the consistency of data scale, the relevanceambiguity, and the data time alignment are all need to befurther taken into account.

IV. CONCLUSIONS AND EXPECTATIONS

According to the investigations mentioned above, it canbe realized that each magnetic survey method has its uniquesuperiority and defect, and each single method has difficultyto meet the requirement of modern geophysical detectionapplications. With the development of the magnetic surveytechnique and instrument system, the future magnetic surveywill change from the traditional work mode, i.e., measuringthe total magnetic field intensity or its gradient to the vectorinformation of the magnetic field, then to multi-parametermeasurements. The magnetic survey application field willchange from the traditional ground magnetic survey methodwith low efficiency and shallow detection depth to highefficient airborne magnetic survey, deep well magnetic survey,and abysmal sea magnetic survey, then gradually to the jointdetection and interpretation combining with the five kindsof magnetic survey methods mentioned above. Through thecombination of a variety of detection methods to realize advan-tageous complementarities, further improvement of detection

accuracy and inversion resolution for Earth’s magnetic fieldwill be implemented.

Since the multi-survey technique fusion is immature, itsuggests that fusion research should move from a feasibilitystudy to problem-solving development. The research shoulddemonstrate fusion as an effective solution in the context of aspecific problem or application rather than simply showinga differently formulated fusion technique applicable to anumber of general problems. To facilitate future research onmulti-survey technique fusion at its initial stage, benchmarkmagnetic data and assessment protocols need to be createdand established to fill the gap for evaluating different fusionmethodologies. This remains a topic for our future work.

ACKNOWLEDGMENT

This work is partly supported by the National NaturalScience Foundation of China under Grant No. 41904164, theNatural Science Foundation of Hubei Province of China underGrant No. 2020CFB610, the Foundation of Wuhan Scienceand Technology Bureau under Grant No. 2019010701011411,the Foundation of National Key Research and DevelopmentProgram of China under Grant No. 2018YFC1503702, andthe Fundamental Research Funds for the Central Universities,China University of Geosciences (Wuhan) under Grant No.CUG190628.

REFERENCES

[1] S. Maus, U. Barckhausen, H. Berkenbosch, N. Bournas, J. Brozena,V. Childers, F. Dostaler, J. Fairhead, C. Finn, R. Von Frese et al.,“Emag2: A 2–arc min resolution earth magnetic anomaly grid com-piled from satellite, airborne, and marine magnetic measurements,”Geochemistry, Geophysics, Geosystems, vol. 10, no. 8, 2009.

[2] H. Liu, X. Wang, J. Bin, H. Dong, J. Ge, Z. Liu, Z. Yuan, J. Zhu, andX. Luan, “Magnetic gradient full-tensor fingerprints for metallic objectsdetection of a security system based on anisotropic magnetoresistancesensor arrays,” AIP Advances, vol. 10, no. 1, p. 015329, 2020.

[3] A. Y. Denisov, V. A. Sapunov, and B. Rubinstein, “Broadband modein proton-precession magnetometers with signal processing regressionmethods,” Measurement Science and Technology, vol. 25, no. 5, p.055103, 2014.

[4] H. Liu, X. Wang, H. Wang, J. Bin, H. Dong, J. Ge, Z. Liu, Z. Yuan,J. Zhu, and X. Luan, “Magneto-inductive magnetic gradient tensorsystem for detection of ferromagnetic objects,” IEEE Magnetics Letters,vol. 11, no. 1, p. 8101205, 2020.

[5] S. Chen, S. Zhang, J. Zhu, and X. Luan, “Accurate measurementof characteristic response for unexploded ordnance with transientelectromagnetic system,” IEEE Transactions on Instrumentation andMeasurement, vol. 69, no. 4, pp. 1728–1736, 2020.

MEASUREMENT 13

[6] S. Chen, H. Jiang, J. Zhu, Y. Qiu, and S. Zhang, “Classificationof unexploded ordnance-like targets with characteristic response intransient electromagnetic sensing,” Journal of Instrumentation, vol. 14,no. 10, p. P10011, 2019.

[7] H. Liu, Z. Liu, S. Liu, Y. Liu, J. Bin, F. Shi, and H. Dong, “A nonlinearregression application via machine learning techniques for geomagneticdata reconstruction processing,” IEEE Transactions on Geoscience andRemote Sensing, vol. 57, no. 1, pp. 128–140, 2019.

[8] C. Tan, J. Wang, and Z. Li, “A frequency measurement methodbased on optimal multi-average for increasing proton magnetometermeasurement precision,” Measurement, vol. 135, pp. 418–423, 2019.

[9] H. Liu, W. Luo, H. Dong, J. Ge, Z. Liu, Z. Yuan, J. Zhu, and H. Zhang,“Design and implementation of a tuning-matching framework for ahigh-sensitivity broad band proton precession magnetometer sensingcoil,” IEEE Sensors Journal, vol. 1, no. 20, pp. 127–134, 2020.

[10] K. Dong, F. Jin, W. Mo, J. Song, and W. Cheng, “Effect of amor-phous/crystalline material doping on the microstructure and magneticproperties of fept thin films,” IEEE Transactions on Magnetics, vol. 53,no. 11, pp. 1–4, 2017.

[11] R. M. Winslow, C. L. Johnson, B. J. Anderson, H. Korth, J. A.Slavin, M. E. Purucker, and S. C. Solomon, “Observations of mercury’snorthern cusp region with messenger’s magnetometer,” GeophysicalResearch Letters, vol. 39, no. 8, p. L08112, 2012.

[12] C. Liao, M. K. Manhard, B. Bilgic, Q. Tian, Q. Fan, S. Han, F. Wang,D. J. Park, T. Witzel, J. Zhong et al., “Phase-matched virtual coilreconstruction for highly accelerated diffusion echo-planar imaging,”NeuroImage, vol. 194, pp. 291–302, 2019.

[13] X. Zeng, M. Yu, and D. Wang, “A new probabilistic multi-hop broad-cast protocol for vehicular networks,” IEEE Transactions on VehicularTechnology, vol. 67, no. 12, pp. 12 165–12 176, 2018.

[14] H. Liu, H. Dong, Z. Liu, J. Ge, B. Bai, and C. Zhang, “Constructionof an Overhauser magnetic gradiometer and the applications in geo-magnetic observation and ferromagnetic target localization,” Journal ofInstrumentation, vol. 12, no. 10, p. T10008, 2017.

[15] H. Liu, H. Dong, Z. Liu, J. Ge, W. Luo, C. Zhang, Z. Yuan, J. Zhu,and H. Zhang, “A comprehensive study on the weak magnetic sensorcharacter of different geometries for proton precession magnetometer,”Journal of Instrumentation, vol. 13, no. 9, p. T09003, 2018.

[16] B. Bai, H. Liu, J. Ge, and H. Dong, “Research on an improved resonantcavity for Overhauser geomagnetic sensor,” IEEE Sensors Journal,vol. 18, no. 7, pp. 2713–2721, 2018.

[17] G. Hoffman, J. Birtwistle, and E. Hill, “The performance of magnetore-sisitive vector magnetometers with optimised conductor and anisotropyaxis angles,” IEEE Transactions on Magnetics, vol. 19, no. 5, pp. 2139–2141, 1983.

[18] P. Vavassori, “Polarization modulation technique for magneto-opticalquantitative vector magnetometry,” Applied Physics Letters, vol. 77,no. 11, pp. 1605–1607, 2000.

[19] K. Mohamadabadi and M. Hillion, “An automated indoor scalar cal-ibration method for three-axis vector magnetometers,” IEEE SensorsJournal, vol. 14, no. 9, pp. 3076–3083, 2014.

[20] A. Canciani and J. Raquet, “Airborne magnetic anomaly navigation,”IEEE Transactions on Aerospace and Electronic Systems, vol. 53, no. 1,pp. 67–80, 2017.

[21] A. Stele, J. W. Fassbinder, J. W. Hartling, J. Bussmann, J. Schmidt, andC. Zielhofer, “Genesis of magnetic anomalies and magnetic propertiesof archaeological sediments in floodplain wetlands of the fossa car-olina,” Archaeological Prospection, vol. 27, no. 2, pp. 169–180, 2020.

[22] Z. Dou, Q. Han, X. Niu, X. Peng, and H. Guo, “An aeromagneticcompensation coefficient-estimating method robust to geomagneticgradient,” IEEE Geoscience and Remote Sensing Letters, vol. 13, no. 5,pp. 611–615, 2016.

[23] P. Calou and M. Munschy, “Airborne magnetic surveying with adrone and determination of the total magnetization of a dipole,” IEEETransactions on Magnetics, vol. 56, no. 6, pp. 1–9, 2020.

[24] R. J. Blakely, Potential theory in gravity and magnetic applications.Cambridge university press, 1996.

[25] L. Hurwitz and J. Nelson, “Proton vector magnetometer,” Journal ofGeophysical Research, vol. 65, no. 6, pp. 1759–1765, 1960.

[26] J. M. Merayo, P. Brauer, F. Primdahl, J. R. Petersen, and O. V. Nielsen,“Scalar calibration of vector magnetometers,” Measurement scienceand technology, vol. 11, no. 2, p. 120, 2000.

[27] J.-T. Jeng, C.-C. Lu, H.-Y. Hsu et al., “Low-noise tunneling-magnetoresistance vector magnetometers with flux chopping tech-nique,” Measurement, vol. 109, pp. 297–303, 2017.

[28] H. Liu, H. Wang, J. Bin, H. Dong, J. Ge, Z. Liu, Z. Yuan, J. Zhu,and X. Luan, “Efficient noise reduction for the free induction decay

signal from a proton precession magnetometer with time-frequencypeak filtering,” Review of Scientific Instruments, vol. 91, no. 4, p.045101, 2020.

[29] H. Liu, H. Dong, Z. Liu, and J. Ge, “Application of Hilbert-Huangdecomposition to reduce noise and characterize for NMR FID signalof proton precession magnetometer,” Instruments and ExperimentalTechniques, vol. 61, no. 1, pp. 55–64, 2018.

[30] S. Liu, X. Hu, T. Liu, J. Feng, W. Gao, and L. Qiu, “Magnetization vec-tor imaging for borehole magnetic data based on magnitude magneticanomaly,” Geophysics, vol. 78, no. 6, pp. D429–D444, 2013.

[31] J.-M. Vouillamoz, M. Descloitres, J. Bernard, P. Fourcassier, andL. Romagny, “Application of integrated magnetic resonance soundingand resistivity methods for borehole implementation. a case study incambodia,” Journal of Applied Geophysics, vol. 50, no. 1-2, pp. 67–81,2002.

[32] L. Fan, X. Kang, Q. Zheng, X. Zhang, X. Liu, C. Chen, and C. Kang,“A fast linear algorithm for magnetic dipole localization using totalmagnetic field gradient,” IEEE Sensors Journal, vol. 18, no. 3, pp.1032–1038, 2017.

[33] J. Ge, W. Luo, H. Dong, H. Liu, H. Wang, W. Wang, Z. Yuan, J. Zhu,and H. Zhang, “Towed Overhauser marine magnetometer for weakmagnetic anomaly detection in severe ocean conditions,” Review ofScientific Instruments, vol. 91, no. 3, p. 035112, 2020.

[34] W. Doll, T. Gamey, L. Beard, D. Bell, and J. Holladay, “Recentadvances in airborne survey technology yield performance approachingground-based surveys,” The Leading Edge, vol. 22, no. 5, pp. 420–425,2003.

[35] E. d. Barros Camara and S. N. P. Guimaraes, “Magnetic airbornesurvey–geophysical flight,” Geoscientific Instrumentation, Methods andData Systems, vol. 5, no. 1, pp. 181–192, 2016.

[36] M. Mayhew, B. Johnson, and R. Langel, “An equivalent source modelof the satellite-altitude magnetic anomaly field over australia,” Earthand Planetary Science Letters, vol. 51, no. 1, pp. 189–198, 1980.

[37] B. IA and Y. SV, “Using satellite magnetic survey data for spatial-temporal modeling of the geomagnetic secular variation,” Journal ofgeomagnetism and geoelectricity, vol. 49, no. 2-3, pp. 207–227, 1997.

[38] P. Ripka, “Review of fluxgate sensors,” Sensors and Actuators A:Physical, vol. 33, no. 3, pp. 129–141, 1992.

[39] P. Ripka, “New directions in fluxgate sensors,” Journal of Magnetismand Magnetic Materials, vol. 215, pp. 735–739, 2000.

[40] H. Liu, H. Dong, J. Ge, B. Bai, Z. Yuan, Z. Zhao, Y. Liu, J. Zhu, andH. Zhang, “Research and develop of the test apparatus for measuringthe excitation frequency of Overhauser magnetometer probe,” ActaElectronica Sinica, vol. 45, no. 2, pp. 1272–1280, 2017.

[41] H. Liu, H. Dong, J. Ge, Z. Liu, Z. Yuan, J. Zhu, and H. Zhang,“Apparatus and method for efficient sampling of critical parametersdemonstrated by monitoring an Overhauser geomagnetic sensor,” Re-view of Scientific Instruments, vol. 89, no. 12, p. 125109, 2018.

[42] A. W. Overhauser, “Polarization of nuclei in metals,” Physical Review,vol. 92, no. 2, p. 411, 1953.

[43] J. Ge, H. Dong, H. Liu, Z. Yuan, H. Dong, Z. Zhao, Y. Liu, J. Zhu,and H. Zhang, “Overhauser geomagnetic sensor based on the dynamicnuclear polarization effect for magnetic prospecting,” Sensors, vol. 16,no. 6, p. 806, 2016.

[44] P. Ripka, “Advances in fluxgate sensors,” Sensors and Actuators A:Physical, vol. 106, no. 1-3, pp. 8–14, 2003.

[45] P. Ripka and M. Janosek, “Advances in magnetic field sensors,” IEEESensors journal, vol. 10, no. 6, pp. 1108–1116, 2010.

[46] H. Dong, H. Liu, J. Ge, Z. Yuan, and Z. Zhao, “A high-precisionfrequency measurement algorithm for FID signal of proton magnetome-ter,” IEEE Transactions on Instrumentation and Measurement, vol. 65,no. 4, pp. 898–904, 2016.

[47] H. Liu, H. Dong, J. Ge, B. Bai, Z. Yuan, and Z. Zhao, “Research ona secondary tuning algorithm based on SVD & STFT for FID signal,”Measurement Science and Technology, vol. 27, no. 10, p. 105006, 2016.

[48] H. Liu, H. Dong, J. Ge, Z. Liu, Z. Yuan, J. Zhu, and H. Zhang,“Efficient performance optimization for the magnetic data readout froma proton precession magnetometer with low rank constraint,” IEEETransactions on Magnetics, vol. 55, no. 8, p. 9300104, 2019.

[49] K. L. Kvamme, “Geophysical surveys as landscape archaeology,”American Antiquity, vol. 68, no. 3, pp. 435–457, 2003.

[50] P. Goldberg, R. D. Mandel, R. Sternberg, A. S. Gilbert, and V. T.Holliday, “Encyclopedia of geoarchaeology (encyclopedia of earthsciences series).” 2017.

[51] H. Liu, Z. Liu, H. Dong, J. Ge, Z. Yuan, J. Zhu, H. Zhang, and X. Zeng,“Recurrent neural network-based approach for sparse geomagnetic data

MEASUREMENT 14

interpolation and reconstruction,” IEEE Access, vol. 7, no. 1, pp.33 173–33 179, 2019.

[52] H. Liu, H. Dong, J. Ge, Z. Liu, Z. Yuan, J. Zhu, and H. Zhang, “Afusion of principal component analysis and singular value decomposi-tion based multivariate denoising algorithm for FID transversal data,”Review of Scientific Instruments, vol. 90, no. 3, p. 035116, 2019.

[53] W. Xiaomei, T. Yuntian, W. Xizhen, W. Chen, and W. Qiong, “The testmethod for the performance of fluxgate magnetometer,” Seismologicaland Geomagnetic Observation and Research, vol. 29, no. 6, pp. 88–94,2008.

[54] X. Wang, Y. Teng, C. Wang, X. Fan, and J. Ma, “Development of fieldarray observation technique system with fluxgate magnetometers,” ActaSeismologica Sinica, vol. 34, no. 3, pp. 389–396, 2012.

[55] M. Prouty, R. Johnson, I. Hrvoic, and A. Vershovskiy, “Geophysicalapplications,” Optical Magnetometry, p. 319, 2013.

[56] V. Korepanov, “Geomagnetic instrumentation for repeat station survey,”in Geomagnetics for Aeronautical Safety. Springer, 2006, pp. 145–166.

[57] A. D. Chave, A. W. Green Jr, R. L. Evans, F. N. Spies, and J. H.Filloux, “Report of a workshop on technical approaches to constructionof a seafloor geomagnetic observatory.” WOODS HOLE OCEANO-GRAPHIC INSTITUTION MA, Tech. Rep., 1995.

[58] S. Linzen, A. Chwala, V. Schultze, M. Schulz, T. Schuler, R. Stolz,N. Bondarenko, and H.-G. Meyer, “A lts-squid system for archaeolog-ical prospection and its practical test in peru,” IEEE Transactions onApplied Superconductivity, vol. 17, no. 2, pp. 750–755, 2007.

[59] H.-G. Meyer, K. Hartung, S. Linzen, M. Schneider, R. Stolz, W. Fried,and S. Hauspurg, “Detection of buried magnetic objects by a squidgradiometer system,” in Detection and Sensing of Mines, ExplosiveObjects, and Obscured Targets XIV, vol. 7303. International Societyfor Optics and Photonics, 2009, p. 73031X.

[60] B. Gavazzi, P. Le Maire, M. Munschy, and A. Dechamp, “Fluxgatevector magnetometers: Compensated multi-sensor devices for ground,uav and airborne magnetic survey for various application in near surfacegeophysics,” EGUGA, p. 19382, 2017.

[61] T. Lin, Y. Zhang, and J. Lin, “Three-component magnetic resonancenoise suppression device and method,” May 28 2014, PRC PatentCN201410088813.0.

[62] J. Lin, Y. Zhang, Y. Yang, Y. Sun, and T. Lin, “Anti-saturationsystem for surface nuclear magnetic resonance in efficient groundwaterdetection,” Review of Scientific Instruments, vol. 88, no. 6, p. 064702,2017.

[63] T. Chen, W. Li, S.-q. Xu, Z. Liang, and X. Wu, “The method ofborehole three-component magnetic survey and its effective originallyexploration,” Progress in Geophysics, vol. 23, no. 3, pp. 892–897, 2008.

[64] J. B. Silva and G. W. Hohmann, “Interpretation of three-componentborehole magnetometer data,” Geophysics, vol. 46, no. 12, pp. 1721–1731, 1981.

[65] R. L. Parker and G. J. Daniell, “Interpretation of borehole magnetome-ter data,” Journal of Geophysical Research: Solid Earth, vol. 84, no.B10, pp. 5467–5479, 1979.

[66] Q. Zhang, W. Al-Nuaimy, and Y. Huang, “Interpretation of boreholemagnetometer data for the detection and characterisation of unexplodedbombs,” Journal of applied geophysics, vol. 61, no. 3-4, pp. 206–216,2007.

[67] K. Bram, J. Draxler, G. Hirschmann, G. Zoth, S. Hiron, and M. Kuhr,“The KTB borehole-germany’s superdeep telescope into the earth’scrust,” Oilfield Review, vol. 7, pp. 4–22, 1988.

[68] Q. Wang, X. Li, and L. Xu, “The high-precision borehole three-component magnetic measurement system: An effective tool in searchfor deep deposits in mines,” Geophys Geochem Explor, vol. 33, no. 3,pp. 235–239, 2009.

[69] R. Emmermann and J. Lauterjung, “The german continental deepdrilling program KTB: overview and major results,” Journal of Geo-physical Research: Solid Earth, vol. 102, no. B8, pp. 18 179–18 201,1997.

[70] W. Bosum, U. Casten, F. Fieberg, I. Heyde, and H. Soffel, “Three-dimensional interpretation of the KTB gravity and magnetic anoma-lies,” Journal of Geophysical Research: Solid Earth, vol. 102, no. B8,pp. 18 307–18 321, 1997.

[71] F. Kuhnke and G. Musmann, “KTB high temperature triaxial magne-tometer: Sci drilling v2, n4, 1991, p166–179,” in International Journalof Rock Mechanics and Mining Sciences & Geomechanics Abstracts,vol. 29, no. 4. Pergamon, 1992, p. 234.

[72] H. Worm, “Rock magnetism and modeling of magnetic boreholeanomalies,” KTB Report, vol. 93, no. 2, pp. 331–333, 1993.

[73] S. Leonardi and H.-J. Kumpel, “Variability of geophysical log dataand the signature of crustal heterogeneities at the KTB,” GeophysicalJournal International, vol. 135, no. 3, pp. 964–974, 1998.

[74] C. Virgil, S. Ehmann, A. Hordt, M. Leven, and E. Steveling, “Re-orientation of three-component borehole magnetic data,” GeophysicalProspecting, vol. 63, no. 1, pp. 225–242, 2015.

[75] S. Ehmann, A. Hordt, M. Leven, and C. Virgil, “Paleomagnetic incli-nation and declination from three-component borehole magnetometerdata-new insights from logging in the louisville seamounts,” Journal ofGeophysical Research: Solid Earth, vol. 120, no. 1, pp. 18–41, 2015.

[76] T. Liu, W. Gao, J. Feng, Y. Xi, and Y. Ou, “Euler deconvolution ofborehole three-component magnetic gradient tensor and its applica-tion,” Geophysical and Geochemical Exploration, vol. 37, no. 4, pp.633–639, 2013.

[77] G. Yuan, B. Ma, and G. Zhang, “New dynamic of borehole geophysicaltechnology application of mineral exploration abroad,” GeologicalScience and Technology Information, vol. 35, no. 1, pp. 184–189, 2016.

[78] J. I. Boyce, E. G. Reinhardt, A. Raban, and M. R. Pozza, “Marinemagnetic survey of a submerged roman harbour, caesarea maritima,israel,” International Journal of Nautical Archaeology, vol. 33, no. 1,pp. 122–136, 2004.

[79] C. Vacquier, N. Steenland, R. Henderson, and I. Zietz, “Interpretationof aeromagnetic maps. memoir 47,” Geol. Soc. Am., Boulder, vol. 47,pp. 1–150, 1951.

[80] B. Yao, “Marine geophysics and its impact on development of marinegeology,” Marine Geology & Quaternary Geology, vol. 4, no. 33, pp.21–27, 2011.

[81] E. Weiss, B. Ginzburg, T. R. Cohen, H. Zafrir, R. Alimi, N. Salomonski,and J. Sharvit, “High resolution marine magnetic survey of shallowwater littoral area,” Sensors, vol. 7, no. 9, pp. 1697–1712, 2007.

[82] F. Zong, R. Deng, L. Ren, J. Gu, P. Zou, and K. Yu, “A marinemagnetometer array system,” Hydrogr Surv Chart, vol. 35, no. 1, pp.45–47, 2015.

[83] H. Dong, S. Hu, J. Ge, H. Liu, W. Luo, Z. Yuan, J. Zhu, and H. Zhang,“A high-precision and fast-sampling frequency measurement methodbased on FPGA carry chain for airborne optically pumped cesiummagnetometer,” Review of Scientific Instruments, vol. 89, no. 7, p.075001, 2018.

[84] N. Seama, “A newly developed deep-tow three component magnetome-ter,” Eos Trans., AGU, vol. 78, p. 46, 1997.

[85] J. S. Gee and S. Cande, “A surface-towed vector magnetometer,”Geophysical research letters, vol. 29, no. 14, pp. 15–1–15–4, 2002.

[86] R. J. Blakely, A. Cox, and E. J. Iufer, “Vector magnetic data fordetecting short polarity intervals in marine magnetic profiles,” Journalof geophysical Research, vol. 78, no. 29, pp. 6977–6983, 1973.

[87] B. C. Horner-Johnson and R. G. Gordon, “Equatorial pacific magneticanomalies identified from vector aeromagnetic data,” GeophysicalJournal International, vol. 155, no. 2, pp. 547–556, 2003.

[88] C. R. German, D. R. Yoerger, M. Jakuba, T. M. Shank, C. H. Langmuir,and K.-I. Nakamura, “Hydrothermal exploration with the autonomousbenthic explorer,” Deep Sea Research Part I: Oceanographic ResearchPapers, vol. 55, no. 2, pp. 203–219, 2008.

[89] D. R. Yoerger, “Autonomous benthic explorer (abe): an auv optimizedfor deep seafloor studies,” in Proc. 7th Int. Symp. on UnmannedUntethered Submersible Technology, Durham, NC, 1991, 1991, pp. 60–70.

[90] C. Tao, T. Wu, C. Liu, H. Li, and J. Zhang, “Fault inference andboundary recognition based on near-bottom magnetic data in the longqihydrothermal field,” Marine Geophysical Research, vol. 38, no. 1-2, pp.17–25, 2017.

[91] M. Dransfield, A. Christensen, and G. Liu, “Airborne vector magneticsmapping of remanently magnetized banded iron formations at rocklea,western australia,” Exploration Geophysics, vol. 34, no. 2, pp. 93–96,2003.

[92] H. Huang, “Airborne geophysical data leveling based on line-to-linecorrelations,” Geophysics, vol. 73, no. 3, pp. F83–F89, 2008.

[93] H. Liu, H. Dong, J. Ge, Z. Liu, Z. Yuan, J. Zhu, and H. Zhang, “High-precision sensor tuning of proton precession magnetometer by combin-ing principal component analysis and singualr value decomposition,”IEEE Sensors Journal, vol. 19, no. 21, pp. 9688–9696, 2019.

[94] H. Liu, J. Bin, H. Dong, J. Ge, Z. Liu, Z. Yuan, J. Zhu, and H. Zhang,“Adaptive pre-whiten filtering for the free induction decay transversalsignal in weak magnetic detection,” Review of Scientific Instruments,vol. 90, no. 10, p. 104502, 2019.

[95] C. Hardwick, “Important design considerations for inboard airbornemagnetic gradiometers,” Exploration Geophysics, vol. 15, no. 4, pp.266–266, 1984.

MEASUREMENT 15

[96] A. Salem, K. Ushijima, T. J. Gamey, and D. Ravat, “Automaticdetection of uxo from airborne magnetic data using a neural network,”Subsurface sensing technologies and applications, vol. 2, no. 3, pp.191–213, 2001.

[97] M. Dransfield, “Searchlights for gravity and magnetics,” Geophysics,vol. 80, no. 1, pp. G27–G34, 2015.

[98] M. Munschy and S. Fleury, “Scalar, vector, tensor magnetic anomalies:measurement or computation?” Geophysical Prospecting, vol. 59, no.Advances in Electromagnetic, Gravity and Magnetic Methods forExploration, pp. 1035–1045, 2011.

[99] J. Ge, H. Li, W. Luo, H. Dong, H. Liu, H. Wang, W. Wang,Z. Yuan, J. Zhu, and H. Zhang, “Aeromagnetic system for a multi-rotorunmanned aerial vehicle based on the overhauser sensor,” Journal ofInstrumentation, vol. 14, no. 01, p. P01015, 2019.

[100] J. Lin, S. Diao, Y. Zhang, Y. Zhang, and J. Zhao, “Research progress ofgeophysical vector magnetic field survey technology,” Chinese ScienceBulletin, vol. 62, no. 23, pp. 2606–2618, 2017.

[101] D. V. Smith and R. E. Bracken, “Field experiments with the tensormagnetic gradiometer system for uxo surveys: a case history,” in SEGTechnical Program Expanded Abstracts 2004. Society of ExplorationGeophysicists, 2004, pp. 806–809.

[102] B. Narod and J. Bennest, “Ring-core fluxgate magnetometers for use asobservatory variometers,” Physics of the Earth and Planetary Interiors,vol. 59, no. 1-2, pp. 23–28, 1990.

[103] Y. Sui, G. Li, S. Wang, and J. Lin, “Compact fluxgate magnetic full-tensor gradiometer with spherical feedback coil,” Review of ScientificInstruments, vol. 85, no. 1, p. 014701, 2014.

[104] S. Keenan, J. Young, C. Foley, and J. Du, “A high-tc flip-chip squidgradiometer for mobile underwater magnetic sensing,” SuperconductorScience and Technology, vol. 23, no. 2, p. 025029, 2010.

[105] T. J. Gamey, W. E. Doll, and L. P. Beard, “Initial design and testing ofa full-tensor airborne squid magnetometer for detection of unexplodedordnance,” in SEG Technical Program Expanded Abstracts 2004.Society of Exploration Geophysicists, 2004, pp. 798–801.

[106] R. Stolz, V. Zakosarenko, M. Schulz, A. Chwala, L. Fritzsch, H.-G.Meyer, and E. Kostlin, “Magnetic full-tensor squid gradiometer systemfor geophysical applications,” The Leading Edge, vol. 25, no. 2, pp.178–180, 2006.

[107] M. Shen, D. Cheng, Z. An, Y. Wang, and J. Zhao, “Geometry structureoptimization of hexagonal pyramidal full tensor magnetic gradientprobe,” IEEE Transactions on Magnetics, vol. 52, no. 9, p. 6100507,2016.

[108] V. Zakosarenko, R. Stolz, L. Fritzsch, H. Meyer, A. Fleischmann,and C. Enns, “SQUID gradiometer for ultra-low temperature magneticmicro-calorimeter,” Superconductor Science and Technology, vol. 16,no. 12, p. 1404, 2003.

[109] N. Olsen, T. Risbo, P. Brauer, J. Merayo, F. Primdahl, and T. Sabaka,“In-flight calibration methods used for the ø rsted mission: Esa sp onspace magnetometer calibration,” ESA SP, vol. 490, 2001.

[110] M. Munschy, D. Boulanger, P. Ulrich, and M. Bouiflane, “Magneticmapping for the detection and characterization of uxo: Use of multi-sensor fluxgate 3-axis magnetometers and methods of interpretation,”Journal of Applied Geophysics, vol. 61, no. 3-4, pp. 168–183, 2007.

[111] B. Gavazzi, P. Le Maire, M. Munschy, and A. Dechamp, “Fluxgatevector magnetometers: A multisensor device for ground, uav, andairborne magnetic surveys,” The Leading Edge, vol. 35, no. 9, pp.795–797, 2016.

[112] P. Le Maire, L. Bertrand, M. Munschy, M. Diraison, and Y. Geraud,“Aerial magnetic mapping with an unmanned aerial vehicle and afluxgate magnetometer: a new method for rapid mapping and upscalingfrom the field to regional scale,” Geophysical Prospecting, vol. 68,no. 7, pp. 2307–2319, 2020.

[113] J. B. Guerard Jr and R. L. Horton, “The management of executivecompensation in large, dynamic firms: A ridge regression estimation,”Communications in Statistics-Theory and Methods, vol. 13, no. 2, pp.183–190, 1984.

[114] X. L. Z. B. X. Mingqiu, “On the software compensation of the field ofview in fourier transform spectroscopy,” Optoelectronics. Laser, no. 1,p. 5, 1997.

[115] F. Ding, T. Chen, and L. Qiu, “Bias compensation based recursive least-squares identification algorithm for miso systems,” IEEE Transactionson Circuits and Systems II: Express Briefs, vol. 53, no. 5, pp. 349–353,2006.

[116] J. H. Sohn, M. Atzeni, L. Zeller, and G. Pioggia, “Characterisationof humidity dependence of a metal oxide semiconductor sensor arrayusing partial least squares,” Sensors and Actuators B: Chemical, vol.131, no. 1, pp. 230–235, 2008.

[117] P. Wu, Q. Zhang, C. Fei, and G. Fang, “Aeromagnetic gradient com-pensation method for helicopter based on∈-support vector regressionalgorithm,” Journal of Applied Remote Sensing, vol. 11, no. 2, p.025012, 2017.

[118] G. Noriega and A. Marszalkowski, “Adaptive techniques and other re-cent developments in aeromagnetic compensation,” First Break, vol. 35,no. 9, 2017.

[119] P. Wu, Q. Zhang, L. Chen, W. Zhu, and G. Fang, “Aeromagnetic com-pensation algorithm based on principal component analysis,” Journalof Sensors, vol. 2018, 2018.

[120] T. Beravs, S. Begus, J. Podobnik, and M. Munih, “Magnetometercalibration using kalman filter covariance matrix for online estimationof magnetic field orientation,” IEEE Transactions on Instrumentationand Measurement, vol. 63, no. 8, pp. 2013–2020, 2014.

[121] Q. Zhang, H. Pang, and C. Wan, “Magnetic interference compensationmethod for geomagnetic field vector measurement,” Measurement,vol. 91, pp. 628–633, 2016.

[122] D. King-Hele, “Progress of the russian earth satellite sputnik 3 (1958δ),” Nature, vol. 182, no. 4647, pp. 1409–1411, 1958.

[123] B. May, “Atmospheric density variations in the 205 km height regionfrom sputnik 3 satellite,” Planetary and Space Science, vol. 9, no. 9,pp. 529–540, 1962.

[124] R. Langel, R. Estes, G. Mead, E. Fabiano, and E. Lancaster, “Initial ge-omagnetic field model from magsat vector data,” Geophysical ResearchLetters, vol. 7, no. 10, pp. 793–796, 1980.

[125] G. Hulot, C. Eymin, B. Langlais, M. Mandea, and N. Olsen, “Small-scale structure of the geodynamo inferred from oersted and magsatsatellite data,” Nature, vol. 416, no. 6881, pp. 620–623, 2002.

[126] D. Duret, J. Leger, M. Frances, J. Bonzom, F. Alcouffe, A. Perret,J. Llorens, and C. Baby, “Performances of the OVH magnetometer forthe Danish Oersted satellite,” IEEE Transactions on Magnetics, vol. 32,no. 5, pp. 4935–4937, 1996.

[127] S. RJ, J. Turner, and W. DE, “Ørsted and magsat scalar anomaly fields,”Earth, planets and space, vol. 52, no. 12, pp. 1213–1225, 2000.

[128] C. Reigber, H. Luhr, and P. Schwintzer, “CHAMP mission status,”Advances in space research, vol. 30, no. 2, pp. 129–134, 2002.

[129] T. Huang, H. Luhr, H. Wang, and C. Xiong, “The relationship ofhigh-latitude thermospheric wind with ionospheric horizontal current,as observed by CHAMP satellite,” Journal of Geophysical Research:Space Physics, vol. 122, no. 12, pp. 12–378, 2017.

[130] Y. Zhou, S. Ma, H. Luhr, C. Xiong, and C. Reigber, “An empiricalrelation to correct storm-time thermospheric mass density modeled bynrlmsise-00 with CHAMP satellite air drag data,” Advances in SpaceResearch, vol. 43, no. 5, pp. 819–828, 2009.

[131] N. Olsen, E. Friis-Christensen, R. Floberghagen, P. Alken, C. D.Beggan, A. Chulliat, E. Doornbos, J. T. Da Encarnacao, B. Hamilton,G. Hulot et al., “The swarm satellite constellation application andresearch facility (scarf) and swarm data products,” Earth, Planets andSpace, vol. 65, no. 11, pp. 1189–1200, 2013.

[132] R. Langel, G. Ousley, J. Berbert, J. Murphy, and M. Settle, “The magsatmission,” Geophysical Research Letters, vol. 9, no. 4, pp. 243–245,1982.

[133] R. Langel, J. Phillips, and R. Horner, “Initial scalar magnetic anomalymap from magsat,” Geophysical Research Letters, vol. 9, no. 4, pp.269–272, 1982.

[134] M. Acuna, C. Scearce, J. Seek, and J. Scheifele, “The magsat vectormagnetometer: A precision fluxgate magnetometer for the measurementof the geomagnetic field,” NASA Technical Memorandum, p. 79656,1978.

[135] R. A. Langel and R. Estes, “The near-earth magnetic field at 1980determined from magsat data,” Journal of Geophysical Research: SolidEarth, vol. 90, no. B3, pp. 2495–2509, 1985.

[136] Z. An, D. Tan, N. ROTANOVA, and V. Golovkov, “Spherical capharmonic analysis of magsat magnetic anomalies over asia,” ActaGeophysical Sinica, vol. 41, no. 2, pp. 168–173, 1998.

[137] Z. An, N. Rotanova, S. Odintsov, and P. Wang, “Spherical cap harmonicmodels of magsat magnetic anomalies over europe and its adjacentregion,” Acta Geophysical Sinica, vol. 41, no. 4, pp. 468–474, 1998.

[138] N. Olsen, L. T. Clausen, T. J. Sabaka, P. Brauer, J. M. Merayo, J. L.Jørgensen, J.-M. Leger, O. V. Nielsen, F. Primdahl, and T. Risbo,“Calibration of the ørsted vector magnetometer,” Earth, planets andspace, vol. 55, no. 1, pp. 11–18, 2003.

[139] T. Neubert, M. Mandea, G. Hulot, R. Von Frese, F. Primdahl, J. L.Jørgensen, E. Friis-Christensen, P. Stauning, N. Olsen, and T. Risbo,“Ørsted satellite captures high-precision geomagnetic field data,” Eos,Transactions American Geophysical Union, vol. 82, no. 7, pp. 81–88,2001.

MEASUREMENT 16

[140] O. V. Nielsen, J. R. Petersen, F. Primdahl, P. Brauer, B. Hernando,A. Fernandez, J. M. G. Merayo, and P. Ripka, “Development, construc-tion and analysis of the ørsted fluxgate magnetometer,” MeasurementScience and Technology, vol. 6, no. 8, pp. 1099–1115, 1995.

[141] N. Olsen, R. Holme, G. Hulot, T. Sabaka, T. Neubert, L. Tøffner-Clausen, F. Primdahl, J. Jørgensen, J.-M. Leger, D. Barraclough et al.,“Ørsted initial field model,” Geophysical Research Letters, vol. 27,no. 22, pp. 3607–3610, 2000.

[142] C. Reigber, H. Luhr, and P. Schwintzer, “Announcement of opportunityfor CHAMP,” GeoForschungsZentrum Potsdam, 2001.

[143] C. Reigber, P. Schwintzer, and H. Luhr, “The CHAMP geopotentialmission,” Boll. Geof. Teor. Appl, vol. 40, pp. 285–289, 1999.

[144] C. Stolle, H. Luhr, M. Rother, and G. Balasis, “Magnetic signatures ofequatorial spread F as observed by the CHAMP satellite,” Journal ofGeophysical Research: Space Physics, vol. 111, no. A2, 2006.

[145] Y.-L. Zhou, H. Luhr, and P. Alken, “The sidebands of the equatorialelectrojet: General characteristic of the westward currents, as deducedfrom CHAMP,” Journal of Geophysical Research: Space Physics, vol.123, no. 2, pp. 1457–1476, 2018.

[146] J. Du, C. Chen, and L. Wang, “Approximate error estimation of globallithospheric magnetic field models in the era of CHAMP satellite,”Progress in Geophysics, vol. 31, no. 4, pp. 1419–1431, 2016.

[147] V. Korepanov and A. Marusenkov, “Fluxgate magnetometers designpeculiarities,” Surveys in geophysics, vol. 33, no. 5, pp. 1059–1079,2012.

[148] C. Reigber, Z. King, R. Konig, and P. Schwintzer, “CHAMP,” in Aminisatellite mission for geopotential and atmospheric research, SpringAGU Meeting, Baltimore, 1996.

[149] F. Colomb, C. Alonso, C. Hofmann, and I. Nollmann, “SAC-C mission,an example of international cooperation,” Advances in Space Research,vol. 34, no. 10, pp. 2194–2199, 2004.

[150] F. Colomb, C. Alonso, C. Hofmann, L. Frulla, I. Nollmann, J. Milovich,J. Kuba, F. Ares, and M. Kalemkarian, “SAC-C mission and themorning constellation,” in IAF abstracts, 34th COSPAR ScientificAssembly, 2002.

[151] R. Colomb, C. Alonso, and I. Nollmann, “SAC-C mission and theinternational am constellation for earth observation,” Acta Astronautica,vol. 52, no. 9-12, pp. 995–1006, 2003.

[152] N. Olsen, H. Luhr, T. J. Sabaka, M. Mandea, M. Rother, L. Tøffner-Clausen, and S. Choi, “CHAOS-a model of the earth’s magnetic fieldderived from CHAMP, ørsted, and SAC-C magnetic satellite data,”Geophysical Journal International, vol. 166, no. 1, pp. 67–75, 2006.

[153] G. A. Hajj, C. Ao, B. Iijima, D. Kuang, E. Kursinski, A. Mannucci,T. Meehan, L. Romans, M. de La Torre Juarez, and T. Yunck, “CHAMPand SAC-C atmospheric occultation results and intercomparisons,”Journal of Geophysical Research: Atmospheres, vol. 109, no. D6, p.D06109, 2004.

[154] A. Kuvshinov and N. Olsen, “A global model of mantle conductivityderived from 5 years of CHAMP, ørsted, and SAC-C magnetic data,”Geophysical Research Letters, vol. 33, no. 18, p. L18301, 2006.

[155] P. Alken and S. Maus, “Spatio-temporal characterization of the equa-torial electrojet from CHAMP, ørsted, and SAC-C satellite magneticmeasurements,” Journal of Geophysical Research: Space Physics, vol.112, no. A9, p. A09305, 2007.

[156] D. Caruso, “CONAE’s satellite missions,” Acta Astronautica, vol. 2,no. 46, pp. 389–395, 2000.

[157] R. Haagmans, R. Bock, and H. Rider, “Swarm: Esa’s magnetic fieldmission,” An ESA Communications Production, vol. BR-302, pp. 1–20,2013.

[158] E. Friis-Christensen, H. Luhr, and G. Hulot, “Swarm: A constellationto study the earth’s magnetic field,” Earth, planets and space, vol. 58,no. 4, pp. 351–358, 2006.

[159] E. Friis-Christensen, H. Luhr, D. Knudsen, and R. Haagmans, “Swarm-an earth observation mission investigating geospace,” Advances inSpace Research, vol. 41, no. 1, pp. 210–216, 2008.

[160] T. J. Sabaka, L. Tøffner-Clausen, and N. Olsen, “Use of the com-prehensive inversion method for swarm satellite data analysis,” Earth,Planets and Space, vol. 65, no. 11, pp. 1201–1222, 2013.

[161] C. Xiong, C. Stolle, and H. Luhr, “The swarm satellite loss of gps signaland its relation to ionospheric plasma irregularities,” Space Weather,vol. 14, no. 8, pp. 563–577, 2016.

[162] J. L. Jørgensen, E. Friis-Christensen, P. Brauer, F. Primdahl, P. S.Jørgensen, T. H. Allin, T. Denver et al., “The swarm magnetometrypackage,” in Small satellites for Earth observation. Springer, 2008,pp. 143–151.

[163] Y. Li and D. W. Oldenburg, “Joint inversion of surface and three-component borehole magnetic data,” Geophysics, vol. 65, no. 2, pp.540–552, 2000.

[164] C. Benech, A. Tabbagh, and G. Desvignes, “Joint inversion of em andmagnetic data for near-surface studies,” Geophysics, vol. 67, no. 6, pp.1729–1739, 2002.