The atmospheric gravitational noise is much larger than the signals from the subsurface (usually).
The noise in that gravimeter you want to buy is mostly atmospheric and thermal radiation field connected.
Calibrate it against the sun and moon first.
Then against 3D near realtime atmospheric models. (Climate and meteorology are just atmospheric, and the magnetosphere is part of the fields (atmosphere) of a planet or body. The atmosphere of earth is intimately mixed with the gravitational, magnetic, electron and ion density, acoustic and electric fields. The velocity and vorticity fields. All these 3D visualizations are just routine ways to see the whole of what is going on. It should be near real time, global and for the whole solar system. The data is almost already there.
Then it would be ready to scan the oceans and earth’s interior. You can do it in other order, but that is from largest to smallest.
This includes gravimeters, seismometers, accelerometers, vibration sensors, infrasound, microbarographs, lightning detection networks, electromagnetic interference networks, GPS, radio telescopes, radar, lidar, magnetometer, magnetotelluric, and a host of others. These are the networks I have been trying to connect together globally. It is for practical reasons – they are all duplicating basic research on signal processing, microvolt and nanovolt amplification, high sampling rate ADCs, large arrays of sensor elements (pixels or islands or nanostructures), clusters of small detectors, and more things in common.
But a missing piece is the local sensor arrays. I think all the sensor networks listed above can be upgraded to work off common core of methods and components, trained workforce and new industries. it will change transportation, education, power generation and distribution, long range planning, chemical and physical manufacturing industries.
The reason I call it “gravitational engineering” is because I see the future where 3D imaging and composition mapping is routine for whole worlds. And where those calibrated images and visualizations are used instead of the open loop simulations and projections we use now.
I have long wanted to use all these fields to control and fabricate new kinds of materials where the bond energies are 1000 to 100 ElectronVolts per bond. Those are the high energy “atomic fuels” and super strong materials needed for space exploration on a solar system scale – and a path to beyond.