Track the "compositional fields"

ASPECT
1

Hello all,
Sometimes I use the “field” to track where material goes with the default “solver parameter”, but the output show the “<0 or >1 composition”.
For example,

   Postprocessing:
     Writing graphical output:  mesh-box2/solution/solution-00009
     RMS, max velocity:         2.97e-05 m/year, 0.000318 m/year
     Compositions min/max/mass: -0.08593/1.201/1.034e+10 // -0.0586/1.059/2.422e+11

In some model, the compositions max could up to 4;
Does this mean that: the computational results use field in this model is wrong? but why the output don’t show the wrong information and cancel the task? Should I use particles-in-cell to track the material instead?

2

Qyunian:
I assume that you initialize the field with values between zero and one? The scheme we use does not guarantee that values remain in that range and as a consequence, you might see values below or above these due to numerical accuracy issues. You should, however, be able to see that the values get closer to these bounds if you refine the mesh. In other words, as the mesh is refined the values you see should remain between zero and one.
Best
W.