In this work, we present a review of the results obtained in the exploitation of percolation theory and the renormalization group method in the study of soil moisture spatial patterns.
In order to capture critical point in soil moisture spatio-temporal dynamics, we developed an algorithm consisting of three steps: (i) dichotomization, i.e., transforming the soil moisture maps into binary maps; (ii) identification of the largest wet cluster; (iii) scaling transformation, i.e., applying an ad hoc implemented coarse-graining procedure to the binary maps.
The methodology was explored by means of several applications on soil moisture data coming from field measurement, remote sensing, and hydrological modelling over a wide range of spatial scales. From the relations between the occupation probability in soil moisture spatial patterns and the normalized size of the largest cluster at different scales, as well as the scaling behaviour, it is possible to argue that also for this physical system the critical point theory applies.
The critical probability seems to be a structural feature of the catchment, being insensitive to the scale of the analysis, as well as to the parameterization of the methodology.