Soil erosion is recognized as a global threat against the sustainability of the natural ecosystem and the environment because of its severe effects in agricultural productivity, damage to infrastructure and pollution of water bodies. Adverse impacts due to human activities resulting in accelerated soil erosion process have been well documented. Much more attention has been given to study the mechanisms associated with the process of soil erosion in the second half of the 20th century. Different mathematical models have been developed to simulate soil erosion processes and incorporate the result in different options of erosion controls. Modelling soil erosion is a complex process that involves numerous parameters. It is for this reason that even highly sophisticated and advanced erosion prediction models like Water Erosion Prediction Project (WEPP) do not incorporate all mechanisms of the soil erosion process. An obvious gap is the satisfactory explanation and incorporation of soil erosion mechanism associated with the initial portion of microchannels where both inter-rill and rill erosion exist. This study attempts to fill this gap through extension of knowledge in the area of soil erosion mechanism, specifically within the initial portions of rill where both splash erosion and erosion due to shear stress exist. Detachment of soil particles from the soil surface depends on the kinetic energy imparted by raindrops. Therefore, it is essential to estimate kinetic energy as accurately as possible to enable study of soil erosion and infiltration mechanisms. Rainfall simulation is widely used to generate rainfall of desired intensities and durations to study soil erosion, infiltration and other dynamic behaviours of soil. Kinetic energy of a rainfall event is often estimated from its intensity. The actual kinetic energy imparted on a soil surface is generally less than the total value of kinetic energy of a rainfall event. This is because of the cushioning effect of the overland flow. Therefore, there is a potential risk of over prediction of splash erosion by an erosion prediction model that does not account for this cushioning effect. In this study, experiments were carried out to estimate the kinetic energy of three different simulated rainfall events produced by three different combinations of pressures and nozzle sizes. The equipment consisted of a multipurpose hydraulic flume, 2m long and 1.4m wide. Five highly sensitive force transducers were mounted on the surface of the flume to measure the impact of raindrops. Different slopes were represented in the experiment by tilting the flume in four different angles from 0 to 15 degrees. Two tipping bucket rain gauges were used to measure rainfall intensity. The nozzles were placed at a height sufficient to produce terminal velocity by the falling rain drops before they hit the flume surface. Overland flow was generated by continuously supplying water to the inlet tank constructed at the upstream of the hydraulic flume. Responses received from the transducers (in the form of voltage) and from the tipping bucket (in the form of pulses) were recorded at regular intervals. Based on this experimental study, a logarithmic energy loss model that accounts for the depth of shallow overland flow, rainfall intensity and bed slope to estimate potential loss of kinetic energy is proposed. Analysis of the results from the study indicated a significant reduction in kinetic energy when the surface flow starts to build up. The analysis also indicated that a significant portion of the energy is lost even though the flow depth is small. This implies that while splash erosion initially contributes to the total amount of soil erosion, most of the erosion after the initial phase is due to the flow induced shear stress. Another important conclusion of this study is that steeper the slope, the lesser the expected overland flow depth and hence more potential for splash erosion and sheet erosion. The Nash Sutcliffe model efficiency statistic of 90% obtained from this study signifies that the model could be used as a useful predictive tool to estimate rainfall kinetic energy loss. The energy loss model developed as a result of this study can be incorporated in process-based soil erosion models to accurately estimate splash erosion and improve the predictive power of these models. In Addition, the model can be used to estimate the critical depth of overland flow when the kinetic energy approaching the soil surface is practically nil. This critical depth can be used to define the transition zone and explicitly define the term “Rill”. The multipurpose hydraulic flume designed and developed for this study can be used for further studies in area of hydraulic and soil erosion research. The methodology developed in this research will be helpful in carrying out further experiments and improve the proposed energy loss model. The potential Future improvements to the model include the followings: i) incorporating the effect of sediment concentration, ii) using wider ranges of intensities, and iii) using an actual soil bed. / Doctor of Philosopy (PhD)

Thesis (M.S. - Hydrology and Water Resources)--University of Arizona. / Includes bibliographical references.

Thesis (D.Phil.(Geography))-University of Pretoria, 2007. / Includes summary. Includes bibliographic references (leaves 185-188).

Thesis (Ph.D. - Hydrology and Water Resources)--University of Arizona. / Includes bibliographical references (leaves 228-234).

Thesis (Ph.D.)--University of Adelaide, Dept. of Civil and Environmental Engineering, 2002. / "May 2002" Includes list of papers published during this study. Errata slip inserted inside back cover of v. 1. Includes bibliographical references (leaves 331-357).

Thesis (M.Sc. Earth Sciences)--University of Waikato, 2008. / Title from PDF cover (viewed February 23, 2009) Includes bibliographical references (p. 72-75)

Thesis (M.Sc. (Natural Resources Management))--Mahidol University, 2006.

Thesis (Ph.D.)--Victoria University of Wellington, 2009. / Includes bibliographical references.

Thesis (M.Sc.)--Victoria University of Wellington, 2009. / Includes bibliographical references.

Thesis (M.S.)--University of Missouri-Columbia, 2004. / Typescript. Includes bibliographical references (leaves 84-88). Also available on the Internet.