Toward a framework for runoff and soil loss prediction using GUEST technology

Abstract

In recent years, a number of physically based models have been developed for soil loss predictions. GUEST is one such model based on fundamental physical principles and the current understanding of water erosion processes. GUEST is mainly used to determine a soil erodibility parameter. To apply the model in a predictive mode, the model is simplified in a physically meaningful manner for flow-driven erosion processes, and 2 essential hydrologic variables are identified, namely total runoff amount and an effective runoff rate. These variables are required to determine soil loss for individual runoff events. A simple water balance model was developed and used to predict runoff amount from rainfall amount. The efficiency of this runoff amount model in prediction was over 90% using field data. A 1-parameter regression model (r2 ~ 0·9) for the effective runoff rate was also established which uses peak rainfall intensity in addition to rainfall and runoff amounts. The prediction of peak rainfall intensity for a given rainfall amount and storm type was also sought. The field data were from Goomboorian, near Gympie, in south-east Queensland and these data were used to test and validate both models. Results overall are satisfactory and the approach adopted is promising. A framework for soil loss prediction is established within which individual parts can be further refined and improved.