Some researchers have highlighted the fact that most of the river classification schemes have not evolved at the same pace as river morphodynamics models have done it. The most prevailing classification scheme of meandering river was proposed by Brice (1975) and is mainly based on observational criteria. Likewise, thermodynamics principles have been applied on geomorphology over a relatively long period of time. Thus, for instance, a strong analogy between meander angle of deflection and the distribution of momentum in gas dynamics has been identified.
Based on the analysis of curvature data from 16 natural meanders (which totals 52 realizations) ranging from class B to class G related to the Brice classification scheme, we propose a two-parameter meandering classification schemen, namely:  the yearly Shannon wavelet based negentropy gradient (ΔSWT), and  a quantitative continuum of the degree of confinement, which is estimated from the dimensonless Frechet distance (δF*) between the meandering centerline curvature and that of the mean center. Our results show that δF* identifies a threshold of ˜650 to discriminate freely from confined rivers; thereby, scales of the second and third degree of confinement are quantified. Likewise, the proxy parameter ΔSWT suggests that there are 4 degrees of meandering morphodynamics which lay in the intervals [10-1-100], [100-101], [101-102], and [102-103]. Our results also suggest that the lower negentropy corresponds to G1 meanders (two phase, bimodal bankfull sinuosity, equiwidth) and class B2 (single phase, wider at bends, no bars). Class G2 (two phase, bimodal bankfull sinuosity, wider at bends with point bars) and class C (single phase wider a bends, no bars) exhibit higher negentropy (single phase wider at bends width point bars). Likewise, the middle-negentropy group is comprised by both confined meanders (B1, single phase and equiwidth channel, and D, single phase, wider at bends with point bars and chutes) and free meanders (class F, two phase underfit, low water sinuosity).
In the light of these results, we believe that our meandering classification scheme has the potential to complement the prevailing observational Brice classification scheme.