Just did some back-of-the-envelope calculations.
1) Total groundwater used for irrigation is estimated at 545 cubic kilometers per year.
http://www.hydrol-earth-syst-sci-discuss.net/7/3977/2010/hessd-7-3977-2010.pdf
2) Couldn’t find a good source but I’ll ballpark 33% of that usage as “fossil” water from deep aquifers that don’t or barely recharge, leaving 180 km3.
3) Total surface area of the ocean is 361,000,000 square kilometers. Tiny fractions of the mined water end up as glaciers, ice sheets, in lakes, or in surface groundwater, but the overwhelming majority ends up in the ocean.
4) The resulting addition ends up raising sea level by half a millimeter/year; this is 2 inches over 100 years.
5) The 100 year average of yearly sea-level rise is 1.8mm/yr.
6) Yes, that can actually be measured.
7) This indicates that fossil water usage could be a significant component of the (tiny) sea level rises that are occuring.
I was looking at the GRACE site and found this:
“Data from the Gravity Recovery and Climate Experiment (GRACE) helps scientists monitor changes in water storage over large areas as illustrated on the front of this lithograph. Gravity changes correspond to mass changes and scientists can isolate the portion of the total mass change caused by the movement of water. Hydrologists go a step further and combine information from GRACE with soil moisture and other data to isolate changes in groundwater storage, allowing them to monitor monthly water storage changes in aquifers.”
From: http://www.csr.utexas.edu/grace/publications/water_litho.pdf
So maybe the GRL article on GRACE is the one. I have had quite a bit of experience with gravity data, so I am interested to know how they arrive at groundwater storage. I know the USGS has done some microgravity experiments to look at groundwater storage in restricted areas.