Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Master → Kenya → GREEN-AMPT BASED MODELLING OF INFILTRATION USING MICROCATCHMENTS IN AN ASAL SOIL

Moi University (2001)

GREEN-AMPT BASED MODELLING OF INFILTRATION USING MICROCATCHMENTS IN AN ASAL SOIL

CHERUIYOT, HENRY KIPTANUI

Titre : GREEN-AMPT BASED MODELLING OF INFILTRATION USING MICROCATCHMENTS IN AN ASAL SOIL

Auteur : CHERUIYOT, HENRY KIPTANUI

Université de soutenance : Moi University

Grade : MASTER OF PHILOSOPHY IN WATER ENGINEERING 2001

Résumé
Infiltration studies were done on an ASAL soil, which is subject to crusting on the soil surface. The study attempted to model the infiltration characteristics of the soil using the Green-Ampt based models, which included the original Green-Ampt (GA), Green-Ampt- Mein-Larson (GAML), Green-Ampt-Mein-Larson modified (GAMLM), and Layered Green-Ampt (Layered GA) models. This involved the determination of the Green-Ampt parameters and subsequently predict the infiltration depths. The layered GA model consider the soil as two layer system (i.e. crust and subcrust) whereas all the other models assume the soil to be one layer system i.e. homogenous soil profile. The research study was carried out on 3 bare and 3 vegetated plots of rectangular shape each covering an area of 20 m2 constructed on a 5 % natural slope in Nasukuta area of West Pokot district, Kenya. Data on rainfall, runoff, soil moisture, hydraulic conductivity and soil suction head were collected. Rainfall was measured using both autographic and standard rain gauges. Runoffwas collected in the plastic buckets located at the lower end of each plot and the volume measured which was converted to the runoff depth. Soil samples were taken both before and after storm event and the soil moisture content determined using the oven drying method. Green-Ampt parameters were determined by measurement, optimisation and texture based regression equations. These parameters were then appropriately incorporated into the GA based models and used to predict the infiltration depths. The falling head and constant head permeameter tests were performed for the measurement of hydraulic conductivity and their values were 1.5 and 11.0 mmlh respectively. An average value of If/ based on tensiometer measurement was found to be 17.5 mm. In the bare plots, the crust conductivity obtained from the Layered GA model was 0.63 mmlh when the crust thickness was assumed to be 5 mm. The subcrust hydraulic conductivity was determined as 10.8 mrn/h, which was close to the value obtained by the constant head permeameter. The crust is therefore about 17 times less permeable than the subcrust. The optimized value of the effective hydraulic conductivity and suction head were 4.8 mmlh and 10.5 mm respectively. In the vegetated plots, the optimised value of effective hydraulic conductivity and suction head were 6 mmlh and 10 mm respectively. In the bare plots the hypothesis of the Layered GA model worked well as revealed by the high value of the R2 (K=0.86) while predicting the infiltration depths. In the vegetated plots, the hypothesis of the Layered GA model showed poor performance (R2=0.14). The performance of the single layer GA models (whereby homogenous soil profile is assumed) with negligible rainfall interception was satisfactory, as was endorsed by the high values of R2 (GA 3, R2=0.84 ; GAML, R2=O.92 ; and GAMLM, K=0.74). The role of crust seems not that well pronounced in the vegetated plots as in the bare plots. The K used herein is based on fitting about 1:1 line and is termed as the coefficient of efficiency. The GA based models are very sensitive to the hydraulic conductivity. It is therefore necessary to obtain accurate values of hydraulic conductivity for satisfactory prediction of infiltration rates and depths. The runoff efficiencies for the vegetated plots ranged from 0 to 26.7 % with a mean of 6.2 % as compared to the bare plots whose runoff efficiency values ranged from 3.5 to 52.9% and a mean of21.6 %. On average therefore, the bare plots had over 3 times more runoff efficiency than the vegetated plots.

Présentation (Research Kenya)

Page publiée le 2 décembre 2013, mise à jour le 22 février 2018