Fig. 1. (a) Schematic diagram of the main science package, where the miniaturized atom sensor is mounted on the portable active vibration isolation platform. (b) Photo of the portable atomic gravimeter running in a noisy lab.

Fig. 2. (a) Schematic overview of the three-dimensional active vibration isolator. (b) Vibration noise in the vertical direction. Red: the noise spectrum measured directly on the lab floor. Blue: the residual vibration noise on the passive isolator. Black: the residual vibration noise on the isolator with active feedback. (c) Long-term performance of the isolator. Inset: vibrational transfer function of the atom sensor.

Fig. 3. Interferometry fringe for $T=82\mathrm{ms}$. It is obtained by 48 drops in 16 s for chirp up and down, respectively. Each black dot is the probability of atoms in the $|1,0\rangle$ state by the averaging of four drops. The error bar represents the statistical error. The purple and red lines are the fitting according to chirp up and down, respectively.

Fig. 4. Top: the gravity acceleration $\mathrm{\Delta }g$ measured by the portable atomic gravimeter between the 29th October and the 8th November 2019. The setup works continuously for more than 10 days in the noisy lab. The two breaks (from 130 h to 143 h and from 192 h to 201 h) are caused by the lasers out of lock. Bottom: the residue achieved from the corresponding gravity signal subtracted by Earth’s tides.

Fig. 5. Allan deviation of the gravity signal corrected for Earth’s tides in the daytime (red) and at night (black). The ${\tau }^{-1/2}$ slopes represent the corresponding averaging expected for white noise.