If you'll notice in the text overlaid in the settlement plot you quote, this plot was for a short window of time (i.e. 5 years verses the 50 year life of the dam) thus 90% of the 100% of the actual settlement data is not shown in this plot. (I would suggest that an inferred conclusion. when missing 90% of 100% of the data, may be premature.)
Given this, there is a clue that may help understand just how much "surface" measured "settlement" translates into an internal effect in the actual dam leakage penetration and volumetric resulting flow.
If you look at the top crest survey data (922ft elev), at 55+00 to 57+00 to 59+00 - there is a residual downward "V" in 6-1971.
IN 1972 and 1973 this "V" balances out or levels out (overall slope angle). What this "V" is is the remainder of one of the two original "differential settlements" from 1967+ early years that is missing (early plot data not available in the referenced report). BUT the "V" reveals the leftover results that caused the massive saturation, of the left side, of the two "twin" flows within the dam leakage noted in prior images (the first Greening Areas of the dam - that no longer exist today) (see image below).
"Back extrapolating" the plot, these twin differential settlement shear areas - located on both sides centered over the toe axis dam hump layer - would be near 3 to 4 inches from an external reference AND at the crest of the dam. The "crest" is located atop the core region (monuments emplaced in downslope edge side near the crest dam road).
Translating "external surface" measurements to actual potential "seam widths" is a complex venture. In this case, when the measurements are on the "top" of the core, is this "closer to the downstream transition zone - or centered exactly on the core?" either way, it is difficult to extrapolate the actual equivalent "opening" width. But the data is useful in that this "residual V" is tied to observable results/effects in documented history leakage. You can see the dramatic effects of this (image below). The internal flows were completely saturating nearly the entire downslope Zone 3 fill. This can only be done by a POSITIVE hydraulic pore pressure such that upward + diagonal flow exists, even if rainfall were contributing, the "upward" pore pressure existence of the Zone 3 fill deeper saturation kept the hillside from percolating the moisture downward (note the clean areas that did percolate properly IF rain present). The leftover "dirt patches" revealed the extent of porting of internal core materials and the positive pore pressures to port the material(s) to the Zone 3 surface. Fortunately, this "twin" differential settlement "closed" by the balancing out of the combined vertical Zone fill - this was facilitated by the fact that these areas were in the central part of the dam and that the "hump" was a temporary densification "fulcrum" that assimilated into the rest of the layer 1 "settlement". Upstream Zone 2 Transition zone silts assisted in the sealing of the core seams.
(side note: this reveals how "irregularities" in the base & abutments of an earth fill dam can be so dangerous - KC Burke provided a link where you can search out dam failures from hard conduit structures that caused "differential settlement" failure of the earthen dams [hydraulic fracturing of the above embankment - fulcrum forces]. If the toe axis "hump" fill were to be rock, Oroville dam would not exist today.)
So this gives a person "actual" reference information as to what results may occur via "external surface" data with respect to the "internal effects" regarding a volumetric flow response.
What makes the current "Green Wet Area" problematic for a continued "differential settlement" is that the left rock abutment will not "balance" out as did the original toe axis "hump" - i.e. the rock hillside wins. Thus the differential settlement will always be present.
It becomes a game of how much transport material may be provided by the upstream Transition zone to continue to "heal" leakage in the core. Keep in mind that this "transport" is from an erosion fill of the core to where the core eroded away more of its plastic seal base. Eventually, the structural integrity of this plastic base becomes compromised as the "silts" fill volume increases. It's a chain reaction. DWR has no idea how much internal waterflow which may be occurring that is "unseen".
Early Dam operation reveals massive internal flows, reaching the Zone 3 surface, depositing "growth medium favorable" internal Core material - forming the first Greening Areas. Settlement (external) back extrapolated to be near 3 inches at dam crest for these twin seam area leakage defects (since collapsed & healed).
