When a slope has already failed, the interesting question is not whether it was stable. Back-analysis inverts the usual calculation: instead of assuming soil strengths and computing a factor of safety, it takes the observed failure geometry, fixes the factor of safety at one at the instant of movement, and solves for the strength that must have been mobilized. It is the most direct evidence available about what the ground actually did at full scale. It is also a model assembled substantially from inference, and it should be read with that in mind.
Why a forward analysis cannot settle it
A conventional stability analysis takes laboratory strengths and design groundwater conditions and reports a factor of safety. Applied after a failure, it usually returns a number comfortably above one — because the inputs describe intact soil and assumed water conditions rather than the conditions that actually existed on the failure surface at the moment it moved.
Back-analysis reverses which quantity is unknown. The failure itself becomes the datum, and the model is calibrated until it reproduces an event that is already known to have occurred.
Locating the surface is most of the work
The analysis is only as good as the geometry fitted to it. The head scarp and tension cracks define the upslope limit, the toe bulge the downslope one, and comparison of pre- and post-failure topography from LiDAR or photogrammetry gives the displaced mass its shape and its volume.
Between those endpoints the surface is inferred. Borings advanced through the slide mass into stable ground below identify the shear zone by disturbed or slickensided material, and inclinometers grouted across it register the depth at which displacement concentrates. Where movement is ongoing, an inclinometer is the single most persuasive line of evidence available.
The premise: a factor of safety of one
At the instant of failure, driving and resisting forces were in balance by definition. Setting the factor of safety to one and solving for strength converts that observation into a number — and that number reflects the field-scale behavior of the entire surface, including fabric, fissures and variability too large to reproduce in a laboratory specimen.
That is the method's genuine advantage. A hand-sized specimen samples a few cubic centimeters of a deposit; the slide sampled the whole surface at once, under the actual drainage and loading conditions of the day it moved, and produced an unambiguous result.
Strength and pore pressure cannot both be recovered
The equation carries two unknowns that matter: the shear strength along the surface, and the pore-water pressure acting on it. One has to be fixed before the other can be solved for, and the choice is not a technicality.
This is where piezometric data earns its cost. If pore pressures at the depth of the failure surface were measured near the time of movement, strength can be back-calculated with real confidence. If they were not, the analyst assumes a groundwater condition, and every back-calculated strength inherits that assumption.
Limit equilibrium and what it assumes
The standard tools divide the sliding mass into vertical slices and enforce equilibrium, differing mainly in how they treat the forces between slices and which equilibrium conditions they satisfy. Methods that satisfy both force and moment equilibrium are generally preferred where the geometry is not simple.
Most analyses are two-dimensional, run on a section chosen as representative. Real slides have width, and the end resistance a two-dimensional model ignores means back-calculated strength from a plane-strain section tends to be conservative — lower than reality — by an amount that depends on the slide's proportions.
Reconciling the model with the laboratory
A back-calculated strength is a hypothesis until it is compared against measured values. Direct shear results under ASTM D3080, triaxial results under ASTM D4767, and residual strengths from large-displacement testing together bracket what the material can plausibly offer.
Agreement between the back-calculated value and residual laboratory strength is a meaningful finding, consistent with movement along a pre-existing surface. A back-calculated strength well below anything the laboratory can produce signals that the model is wrong somewhere, most often in the assumed water condition.
The honest limits
Back-analysis is non-unique. Different combinations of strength, pore pressure and surface geometry can reproduce the same failure equally well, and the method cannot distinguish among them on its own without independent measurements.
It also returns an average. A single mobilized strength is assigned to a surface that may cross several materials, and the result cannot capture progressive failure, in which strength is exceeded at one point and the surface propagates rather than mobilizing everywhere at once.
How a defensible back-analysis reads
The credible version states its assumed geometry, its assumed piezometric surface and the method used, then presents a sensitivity study showing how the back-calculated strength shifts across the plausible range of each assumption.
Where the result holds steady across that range, it is strong evidence. Where a modest change in assumed groundwater swings the answer across the range of plausible materials, the analysis has identified what was never measured rather than what the soil was, and the report should say so.
This article is general technical orientation, not a failure analysis, an engineering opinion, or advice on any specific matter. Determining the cause of a particular incident requires hands-on examination by a credentialed expert.