Dwivedy, Professor, Department of Mechanical Engineering, Indian Institute of Technology (IIT) Guwahati, for their valuable suggestions, eternal encouragements and timely assistance provided at various stages of research and composition of the thesis. Amit Dubey, Research Scholar from the Department of Civil Engineering, IIT Guwahati, for his help during the preparation of the thesis.
Overview 1
Therefore, both the occurrence of shallow water overland flows and subsurface storm flows along preferred paths are considered dominant and critical hydrological processes in the region. Negi (2001) emphasized the need for micro (area scale) and meso (catchment scale) studies in Himalayan mountain regions to better understand hydrological processes.
Objectives 3
Water infiltration and soil macroporosity 8
The presence of macropores generally leads to heterogeneity of the flow in the soil (Ligon et al., 1977; Germann, 1981). In macroporous soils, plot-scale measurements should be preferred over point measurements (Sarkar et al., 2008a).
Subsurface Water Flow through Soil Macropores 10
Experiments with dyes and tracers are popularly used to study water flow through soil macropores (Kissel et al., 1973). Larsbo and Jarvis (2003) developed a physically based, one-dimensional, numerical model for water and solute transport in macroporous soils. .
Overland Flow on Hillslopes 13
The length and slope of the slope also influenced the process of runoff formation and other hydraulic characteristics of the overland flow. Experimental observations of runoff generation and flow concentration on irregular slopes indicated that the topography of the sloping surface controlled the direction and streamlines of overland flow.
Subsurface Water Movement 15
- Soil moisture storage and its distribution on hillslopes 16
- Characterization of soil macropores using dye and tracers 18
- Subsurface stormflow 20
- Lateral preferential flow 27
The spatial and temporal scales of measurements are often a serious drawback considering the wide spatio-temporal variation of the preferred flow characteristics. One of the two mounds showed higher subsurface flow response due to the presence of soil micropores and mesopores developed by plant roots.
Conclusion 31
This chapter presents a description of the study area and the selected experimental hillslope in the Brahmaputra river basin. A detailed description of the experimental setup and methods for the in situ experiments is given herein.
Soil, Vegetation, and Climate 36
Design of the Experimental Setup 37
Limitations of the sheet flow generation system 40
Measurement of Infiltration 43
Double ring infiltrometer tests 43
Infiltrations were measured at 10 plot locations (Fig. 3.9) under dry (unsaturated) and wet (saturated or at field capacity) antecedent conditions (Table 3.4). The following linear function was found to represent quite reasonably the variation of uniform infiltration rates along the length of the plot for wet antecedent conditions and dense vegetation (Figure 3.12).
Estimation of infiltration from runoff plot experiments 49
Furthermore, it can be observed that the saturated hydraulic conductivity (Table 3.1) of the soil matrix, as calculated from soil texture data (Schaap et al., 1998; Schaap, 1999), is much lower than the calculated spatially averaged uniform preferential infiltration rate. from runoff plot experiments. This clearly indicates the high variability of the infiltration rate due to the presence of macropores in the slope soil.
Characterization of Soil Macropores 51
Experimental setup and methodology 52
A one meter long and 2.5 cm diameter profile probe soil moisture meter (Delta T), equipped with a handheld data logger (Delta T), was placed through the middle of the soil column to the bottom of the hollow cylinder for monitoring the volumetric moisture content of ground. By this time the profiler readings indicated a steady saturated condition of the soil column. This consideration is reasonable since the frequent high intensity storm events of the region can lead to such depth of ponded water on the ground surface.
Image analyses 55
- Horizontal dye pattern analysis 56
- Vertical dye pattern analysis 58
The analyzes of the horizontal and vertical dye patterns of the images were performed as follows (Weiler, 2001; Weiler and Flühler, 2004). The percentage of dye coverage areas of horizontal images was used to calculate the statistical parameters of the vertical dye patterns. For a given section, volume density was calculated as the total number of colored paths divided by the width of the section.
Subsurface Soil Moisture Monitoring System 63
Calibration of the profile probe soil moisture meter 63
The measured sensor output (mV) can be directly related to the square root of the apparent dielectric constant (√ε) of the soil. The wet weight (Ww gm) and volume (Vv cm3) of the soil were recorded before drying the samples in an oven. For oven-dried samples (ie θ = 0), the sensor outputs (V0) were observed and finally the dry weight (W0 gm) of the samples was taken.
Study of soil moisture profile in the hillslope plot 66
Interesting observations can be made from the soil moisture distribution patterns in the hilly soil. 3.25(a-d) Temporal variations in soil moisture content at different depths (b.g.l.) in the hill slope diagram during and immediately after the runoff experiment. 3.26(a-d) Temporal variations in soil moisture content at different depths (b.g.l.) in the hill slope diagram for longer durations after the runoff experiment.
Observation of Overland Flow in the Hillslope Plot 73
In situ overland flow experiments 73
- Experimental results of overland flow 74
Overland flow experiments were conducted in three different seasons representing three distinct vegetation conditions to study the effect of varying plot physical condition on infiltration and overland flow behavior. 3.28(a-f) shows the typical nature of the output hydrographs obtained from experiments conducted under different conditions. This phase represents a constant and steady state of soil infiltration, referred to as the preferential infiltration rate (fb).
Observation of overland flow under natural storm events 80
- Observational setup and instruments 80
- Runoff generation from the hillslope plot 81
- Evaluation of the inflow intensity-infiltration
Only four of these storms resulted in surface runoff from the slope face (Table 3.12). For storm events in 2010, a scatter plot was prepared between the precipitation depth of storm events and the 7-day API (Figure 3.32). Non-runoff storm events are well separated from runoff events in the scatterplot.
Experimental Evidences of Lateral Subsurface Flow 88
Water table measurements in all piezometers were carried out before, during and after the completion of the discharge experiments. Very little lateral spread of water was visible on the right side of the plot. 3.41(a-e) Temporal variation of water table depths above impermeable layer observed at different piezometer locations for experiment E-12.
Summary of the Chapter 93
Description of the model 97
Overland flow experiments carried out on a slope plot showed that the constant preferred infiltration rate (fb) depends on the inflow intensity (i) and the degree of vegetation. To simulate overland flow behavior on a slope face for a given runoff experiment, the measured inflow intensity and constant preferred infiltration rate were given as inputs. Overland flow model input i) Inflow intensity (i) . ii) Constant preferred rate of infiltration (fb). i) Observed runoff coefficient ii) Measured runoff hydrograph.
Overland flow model simulations 101
4.2(a-f) Observed and overland flow model simulated outflow hydrographs from the hillslope plot for some of the trials. 4.3(a-c) Variation of simulated overland flow depths for different inflow intensities for moderate vegetation condition of the plot. For low overland flow depth, variation of the surface roughness on a vegetated hillslope is mainly characterized by degree of vegetation.
Modeling Lateral Subsurface Stormflow 107
- Physical concept of the model 108
- Mathematical formulation of subsurface stormflow 110
- Computation of macropore flow 111
- Numerical solution of subsurface stormflow 112
- Simulations of the subsurface stormflow model 116
- Design of soil macropore structure through simulations 117
Thus, the total lateral flow per unit length of the hydrologically active macropores at a given time can be calculated as. It is clear from the figure that, regardless of the number of soil layers, the temporal variations in the groundwater profile were largely dependent on the hydrologically active macroporosity of the soil. Similar results were obtained from the simulations of the runoff experiments performed on the hill slope diagram.
Summary of the Chapter 127
The model's simulated outflow hydrograph was then compared to the measured hydrograph at the downstream end of the plot. The distribution of the macropores in different soil layers did not appear to have a significant effect. In most cases, a large part of the runoff occurred underground within 2-3 hours of cessation of recharge.
Overland Flow Behavior on Hillslopes 137
Effect of relief on overland flow 142
To study the effect of relief on overland flow hydrographs, keeping other conditions equal, the relief of the three types of hillslopes was changed from 20 m to 10 m, and the resulting Digital Elevation Models (DEM) were generated. Comparison between the runoff hydrographs revealed (Table 5.5) that stable peak discharge conditions were achieved in all three types of hillslopes. However, due to flattening of the slope, the values of tp and tr were increased.
Overland flow in non-macroporous soil 142
For a sandy loam soil on a slope, the Green-Ampta infiltration parameters were taken as: suction height of the soil wet front (ψ). As expected, the magnitude of peak flow was much greater (Table 5.6) than for slopes with preferential infiltration. Most importantly, it was evident from the maximum depth of propagation of the saturated water front below the ground surface (Table 5.6) that in no case did the water front reach the impervious layer to initiate any active subsurface storm flow.
Subsurface Stormflow from Hillslopes 146
Effect of relief on subsurface stormflow 150
The subsurface storm flow response of the three theoretical slopes was analyzed based on the model simulations. To characterize the macroporosity of the hilly soil, undisturbed soil columns were collected from the plot. The model was used (1) to capture the rapid subsurface storm surge that is clearly visible on the hillslope, (2) to represent the contributions of the soil matrix and macropores in the rapid interflow that occurs immediately after the high intensity runoff events, and (3) to provide a surrogate. indicator of hydraulically effective lateral macroporosity of the slope.
Effect of changing lateral macroporosity on 151
Major Conclusions 169
Due to the presence of high macroporosity in the soil, point measurements of infiltration showed wide spatial variations within the hillslope plot. Observation of overland flow response under natural storm events revealed that overland flow generation from the hillslope was a threshold-dependent process. The subsurface storm flow model was simulated to optimize the active soil macropore structures of the hillslope plot under different rates of recharge.
Scopes for Future Research 172
By including a vertical macropore flow component, the overland flow and the subsurface storm flow model can be coupled. Better instruments can be used to provide more accurate quantitative data on soil macropore structures, macropore flow rates, subsurface geological formations, etc. Finally, the simplifying assumptions adopted in model development can be improved by incorporating a more realistic physical definition of complex flow processes.
Experimental Investigations
3.5 (a-b) shows the intensity, frequency and duration curves of two rainfall stations (Lakhimpur and Guwahati) in the region. 3.25 (a-d) shows temporal variations of volumetric soil moisture content at different depths at four selected locations (P1, P2, P3 and P4) of the plot. The analysis of the measured data showed that the generation of overland flow from the slope surface is a threshold-dependent process.
