FIELD EVALUATION OF INFILTRATION MODELS

Infiltration has a great importance in the watershed management and prediction of flood. Infiltration is defined as a physical phenomenon, in which water penetrates into the soil from surface sources such as precipitation, snowfall, irrigation etc. Information of infiltration is necessary in hydrologic design, watershed management, irrigation, and agriculture. It is, therefore, necessary to have a detailed understanding of infiltration characteristics for a given land use complex. Infiltration is a vital component process of the hydrologic cycle. It is one of the main abstractions accounted for in the rainfall-runoff modeling. In the hydrological process, infiltration divids the water into two parts surface flow and groundwater flow. Soils of different types have different infiltration characteristics. Infiltration rates are affected by a number of factors of which antecedent soil moisture texture of the soil, density and behaviour of the soil. Knowledge of infiltration is essential for any beneficial durable study of hydrological evaluations. In this investigation, the performance of the various infiltration models (Mezencev, Philip’s, Horton’s, Kostiakov, Modified Kostiakov and Lewis and Milne) was evaluated by using double ring infiltrometer on five different locations in National Institute of Technology, Kurukshetra. The aim was to study the ability of the models in accurately predicted cumulative infiltration. The performance of various models was evaluated using evaluation parameter Sum of Squared Error (SSE), Model Efficiency and Root Mean Square Error (RMSE) criteria. The results show that Modified Kostiakov model and Mezencev model are most efficient models with SSE, Model Efficiency and RMSE that are 2. 352, 99.621, 0.400 and 2.483, 99.619, 0.491 (average values) respectively. Hence, Modified Kostiakov and Mezencev model could be used successfully to evaluate the cumulative infiltration of soil for the study area. Keywords: cumulative infiltration, prediction of flood, root means square error, sum of square error