|Ph.D Student||Nezlobin David|
|Subject||Modified Percolation Approaches to Connectivity|
Effects in Runoff Generation
|Department||Department of Civil and Environmental Engineering||Supervisors||Professor Hillel Rubin|
|Mr. Gideon Sinai (Deceased)|
|Full Thesis text|
Description of overland flow connectivity represents one of the most challenging tasks of the modern hydrology; however, there is a clear lack of powerful quantitative approaches to the subject. Percolation theory describes critical changes in connectivity of statistically large partly connected systems. In the current research we apply modified percolation theory approaches for characterizing two different phenomena: first, we study experimentally and theoretically surface runoff connectivity on planar patchy hillslopes. Second, we observe experimentally runoff initiation from raindrops impacting inclined impervious surface, focusing primarily on the critical changes in flow connectivity.
The major findings regarding the surface runoff on patchy hillslopes are as follows:
The classical percolation theory provides a powerful apparatus for describing the runoff connectivity on patchy hillslopes covered by generating and obstructing covers (e.g. bare soil and vegetation), but it ignores strong effect of the overland flow directionality. To overcome this and other difficulties in the percolation theory application, directed approaches can be considered, such as straight percolation (for the planar slopes), quasi-straight percolation and models with limited obstruction. The dependences of the straight percolation quantities on the percentage cover and scale are calculated by Monte-Carlo simulations, and explained in the framework of the combinatorial model of urns with restricted occupancy. The contributing area grows sharply when the runoff generating percentage cover approaches the straight percolation threshold. The character of the contributing area expansion during heavy rainfalls is shown to be significantly dependent on the percentage cover and scale. The nature of the cover-related runoff scale decrease is combinatorial - the probability for the generated runoff to avoid obstruction decreases with scale for to the given percentage cover value. The magnitude of the scale effect is found to be a skewed non-monotonous function of the percentage cover.
In the second part of the research the runoff initiation on planar impervious surfaces has been investigated in laboratory rain simulator experiments using visual and FLIR observations. Depending on the surface inclination several different morphological regimes of flow initiation have been observed - broad flows caused by critical coalescence (very small inclinations), rivulets (moderate inclinations) and flow threads (steep inclinations) with or without sliding. In particular, critical coalescence of drop clusters on the surface leads to the abrupt emergence of surface runoff on impervious surfaces with low inclination, e.g. such as most paved surfaces in urban areas. The role of several non-trivial effects is emphasized, such as coalescence shrink and the effect of wetted trails.