please have a look on the attached documents. it involved 250 articles recently published in the same topic. hope this help

I am not sure, what type of subsurface data you want to correlate with remote sensing techniques. However there are various proxies that can be used to correlate elevation, terrain classes, geology to the surface information. But these are not global correlation, for example, you have to first see if any such correlation is available for local conditions, then you may use , but in absence of that, you have to develop correlation for DEM, elevation, geology with subsurface data (for few sample measurements), then you may apply that correlation to overall local ground.

this is 1 of the example,

https://pubs.geoscienceworld.org/ssa/bssa/article-abstract/107/1/117/350980/proxy-based-vs30-estimation-in-central-and-eastern

With the complex geological structures of earth...how this would be possible.. there should be high rate of aproximation.. or so many shp data layers from boteholes to help .. as in geospatial analysis the problem is how much data you will provide.. but in remote sensing it is just some surface data which could be greatly different not only in depth but in surface ..

Data is available in Grid pattern (80 Points spread over approximately 23000 km2). Available data include

Global Coordinates, Elevation, Soil Classification, SPT N Values, Moisture Content, at 5 ft. 10. ft. and 15 ft. depth. Geological Map of the Area.