Optical system is another vital subsystem in atom interferometer, which provides laser beams with appropriate frequency and power. This subsystem in lab always occupies large space and is not easy to move. The most direct way to miniaturize an optical system is to replace the bulky and heavy optical mounts with miniaturized one.^[68] However, the free-space system is sensitive to external environment such as the vibration noise and the temperature fluctuation. By contrast, fiber optics based on the C-band and second harmonic generation of mature telecom is an alternative method to obtain a stable and compact optical system.^[69,70] This system usually needs only one seed laser and few fiber components, which greatly simplifies the system and is insensitive to misalignment.

A typical fiber system is shown in Fig. 6. The output of a 1560- nm seed laser is divided into two parts. One part is used to lock the frequency of the seed laser, and the other part passes through a phase modulator (PM) to generate the second laser with a different frequency. The two laser beams are amplified by a erbium-doped fiber amplifier (EDFA) and then their frequencies are doubled by a periodically poled lithium niobate crystal-wave guide (PPLN-WG). The microwave source driving the PM can be changed in different steps, and all the required frequencies can be achieved. In addition, an all-fiber system based on 780-nm seed laser was also demonstrated by Fang et al.^[71]

The acceleration measured by the atom gravimeter is the acceleration between atom cloud and reference Raman mirror, which consists of two parts: the gravity acceleration and the vibration of the Raman mirror. These two parts cannot be distinguished experimentally according to the Equivalence principle. To reduce the influence from vibration, researchers usually placed Raman mirror on an isolation platform.^[29,72–74] But this method needs complicated vibration isolator, which is not convenient to transport in practical applications. An alternative way is using mechanical sensors to record vibration data and then compensate this extra vibration phase by combining the corresponding transfer function. The seismometer,^[75] the force balance accelerometer^[76–78] and the opto-mechanical resonator^[79] could be ideal candidates for mechanical sensors. All of these sensors have shown good performance in either improving short-term sensitivity or reconstructing the fringes, which are washed out by harsh vibration noise.

The accuracy is a core indicator to assess the performance of an absolute measurement instrument. Some private^[29,80] and international comparisons were held between absolute gravimeters to find possible systematic errors. International comparison of absolute gravimeters (ICAG) is an official and international comparison, which is held every four years from 1981 and, till now, 10 sessions have been successfully held. In 2009, a transportable atomic absolute gravimeter developed by SYRTE participated in ICAG.^[81] This is the first time that an absolute gravimeter employs atomic ensemble as test mass rather than the macroscopic corner-cube only. In 2017, six atom gravimeters participated in the 10 th-ICAG held in Beijing and exhibited well.^[82] The uncertainty of the best atom gravimeter was 3 μGal, which was developed by HUST.^[44] The gravity measured by two kinds of gravimeters showed a good agreement with each other, which directly demonstrated that atom gravimeter have the same outstanding performance in gravity measurement as commercial gravimeters, which laid a solid foundation for atom gravimeters to give full play in future field application.

Resource exploration and other field constructions also need to measure gravity in a large area while traditional measurement requires lots of manpower and time. A movable and reliable absolute gravimeter can meet this requirement without regular calibration. The movable absolute atom gravimeter developed by HUST achieved a sensitivity of 53$\,{\rm{\mu Gal}}/\sqrt{{\rm{Hz}}}$ and its resolution can reach 1 μGal in an interrogation time of 3000 s. This atom gravimeter can be used for resource exploration.^[70] Besides, a more compact atom gravimeter is under construction for field application. The researchers in UC Berkeley have developed an atom gravimeter which can be transported by a vehicle site to site.^[42] In the laboratory, the sensitivity of this gravimeter could reach 37$\,{\rm{\mu Gal}}/\sqrt{{\rm{Hz}}}$ and the stability was better than 2 μGal when T = 130 ms. To demonstrate field application capabilities, they measured gravity at six locations in Berkeley Hills with a span of 7.6 km and an elevation change of nearly 400 m (see Fig. 7). Using the measured data, the average density of the hills was calculated to be 2.0(2) g/cm³, which was in good agreement with the data collected in 1998. Besides, Lin’s group also integrated their compact atom gravimeter on a truck and measured gravity on both a flat road and a slope road.^[83]

The ocean occupies 70% of the Earth’s area. A more detailed understanding of the structure of the ocean’s gravity field can help to accurately study the shape and internal structure of the Earth. Surface ship based marine gravimetry is one of the high precision methods in marine gravity survey. Bidel and his coworkers introduced an absolute marine gravimeter which has been successfully implemented in real sea condition.^[76] As shown in Fig. 8, the whole gravimeter was placed on a gyro-stabilized platform on ship to ensure the sensitive axis remained vertical at all times. To test their gravimeter, three different strategies were applied. Firstly, they measured the gravity along a known calibration profile, the reproducibility was < 0.5 mGal and the difference of mean value between the test data and known data was < 0.6 mGal. Last two measurements were measuring regional gravity along a grid or circles and inverting the local gravity model. The result not only showed that this atom gravimeter for marine gravimetry can derive a more accurate gravity model than satellites, but also demonstrated that the satellites altimetry may have an offset. For comparison, a relative gravimeter was placed next to atom gravimeter in first measurement and the result showed that atom gravimeter had a better performance in terms of repeatability and stability.

After this successful application, Bidel and his coworkers used the gravimeter to achieve a quantum absolute airborne gravimetry in Iceland with a spatial resolution of 10.5 km,^[84] as shown in Fig. 9. The research strategy was the same as the one on ship. In the first step, they measured gravity back and forth along a straight line between two locations, which showed that repeatability of gravimeter was better than 3.9 mGal. In the second step, they also measured the gravity disturbance in another area where ground gravity data can be obtained. Compared with the results of the two models, the standard deviations and the mean differences were smaller than 6.2 mGal and –1.9 mGal, respectively.

Another atom interferometry experiment based on plane is the ICE project,^[85] which dedicates to test the universality of free-fall (UFF) in microgravity environment. This project plans to simultaneously perform two atom interferometers based on ⁸⁷Rb and ³⁹K to test the UFF at 10⁻¹¹ level by plane and 10⁻¹⁵ for space mission in future. In 2009, they obtained the Ramsey fringes based on ⁸⁷Rb with an interrogation time of 75 ms, which was much longer than T = 20 ms under standard gravity using same setup.^[86] Two years later, they conducted an atom interferometry with MZ configuration on plane.^[87] In this experiment, the interferometry was performed using only ⁸⁷Rb atoms with a temperature of 10 μK. A sensitivity level of $2\times {10}^{-4}\,({\rm{m}}/{{\rm{s}}}^{2})/\sqrt{{\rm{Hz}}}$ was achieved in 0 g, which was 300 times weaker than the vibration of plane. An improved sensitivity of 3.4 × 10⁻⁵ g / shot was obtained during 0 g flight in 2016,^[88] and the interferometry fringe recorded was shown in Fig. 10.

Experiments in space exhibit abundant physical phenomena, which attract more and more researchers’ attention. In recent years, a lot of cold atom experiments,^[89,90] especially the space projects based on cold atom interferometry^[91–95] are proposed. Recently, the BEC (Bose–Einstein condensation) was observed in the international space station,^[96] which offered a reliable platform in space. It is foreseeable that the performance of the atomic interferometer will be greatly improved.

As we all know, the cost of a launch mission is high. To meet the requirement, the Center of Applied Space Technology and Microgravity (ZARM) in the University of Bremen constructed a drop tower, which can provide a weightlessness environment for various space experiments. An outstanding performance of this tower is a relative low residual acceleration below 10⁻⁶ g, which is much less than 0- g platforms. The total height is 141 m and the maximum free-fall time allowed is 9.4 s.^[97] In 2007, a joint team based on the project QUANTUS successfully realized BEC in microgravity in ZARM.^[98] Seven years later, they demonstrated atom interferometry with an improved BEC source, whose temperature drops to 1 nK, thus obtaining a longer expansion time.^[99] In the interferometry, an asymmetric Mach–Zender interferometer was applied. Even if the total time was larger to 0.5 s, the contrast was obtained to be greater than 40% (see Fig. 11).

The cold atom platform, which is compatible with sounding rockets, is another good choice. It extends the time for free evolution to about 6 minutes. In addition, the entire process of sounding rocket can simulate the environment like the one on the satellite as realistically as possible, which will lay a firm foundation for future space projects. Some groundbreaking experiments on atomic optics have been carried out on rockets.^[100,101] Both of them demonstrated the validity of the autonomous frequency stabilization on rocket, which is not easy to accomplish under the condition where peak acceleration is up to 12.1 g and vibration is 8.1 g_rms. On 23 January 2017, the mission MAIUS-1 was launched and a series of experiments were accomplished.^[102] They studied the phase transition from thermal ensemble to a BEC and applied 110 experiments to atom interferometry. Although there are few public reports on atom interferometry, this is the first time that a sounding rocket has been used to demonstrate atom interferometry in microgravity.