Residuals and unresolvable conflicts between different input data are inevitable given the scope of ice-sheet-scale reconstructions. Clusters of conformable observations, for example regarding rel ative age, are generally to be trusted, and given preference over individual observations. One of the crucial points is the data reduction: can we know for certain that, for example, spatially separated patches of glacial lineations that indicate roughly conformable ice-sheet patterns really pertain to the same event? Caution has to be exercised, because there is a significant chance that roughly similar flow patterns are repeated. Iteration is necessary; if conflicts arise it is necessary to backtrack and test another slightly different data reduction.
Was this article helpful?