Title: Modeling The ‘Second’ Soil Erosion On Murram Roads In Uganda

Twaibu Semwogerere

Keywords: roads, road safety, engineering structures, soil erosion, murram, murram roads, prediction models, erosion effect, mathematical modelling, road maintenance, maintenance effects, rills, dunes, soil deposition
Issue Date: March 9, 2016
Abstract: Engineering structures like roads have had failures that have basically relied on our ability to design safe and economic foundations. Erosion, as one of the catastrophic events, has features like rills or dunes on murram/dirt/gravel roads that create a threat to the road users, constructors, and maintainers. These result in increased accident risks, raised costs, and a remarkable effect on the agricultural economies of developing countries like Uganda. Therefore modeling such events has been one of the ways that has been applied to curb such catastrophes. Although soil erosion prediction models contribute to the scientific understanding of erosion processes, most of them are agricultural-based. Mathematical modeling of erosion/deposition on these roads is a unique area of study that needed attention. In this work, a model was constructed to describe erosion from these features formed after some (‘first’) erosion in order to help in their maintenance.

The model was built on three factors using numerical methods: the nature or make of the road, erosion effect, and the repairs or maintenance effects. These factors were also parametized further with the help of a mathematical background and earlier built formulas like Mann’s. In this formulation, the rill/dune effects on the road were assumed to be proportional to the repairs that are necessary to neutralize them. The repairs and the rill/dune sub-models were constructed and analyzed first, and combined numerically in order to construct the general erosion model. It was analyzed with several numerical examples using data from the various regions of Uganda. Comparative and experimental designs were used to generate data and in the models’ formulation. Field and laboratory experiments involved among others measuring runoff speeds of sediment and measuring rill/dune widths/heights. Laboratory experiments involved sieve analysis and liquid limit analysis among others.

It was found out that the rill/dune sub-models could predict the rill/dune effects in terms of specific area or volume of a particular road. Similarly, a repairs sub-model could approximately give the time required for road repairs given a remarkable effect. The erosion model could also ably predict the effects of erosion given proportionate repairs. The model evaluation showed that the effects tremendously reduce given some proportionate repairs. The model was therefore recommended for use by stake holders like UNRA, and Uganda Road Fund. The model shall also help in operating a safe and efficient national road network through responding to the national economic needs, road safety and environmental sustainability.

Presented at: Own PhD Defense


Attachment Name Attachment Type

Modeling The ‘Second’ Soil Erosion On Murram Roads In Uganda