Flood modelling of Periyar basin using SWAT

FLOOD MODELING OF PERIYAR BASIN USING SWAT
PROJECT REPORT
Department of Civil Engineering
T.K.M. College of Engineering, Kollam- 691005
A. P. J. ABDUL KALAM TECHNOLOGICAL UNIVERSITY
2021

FLOOD MODELING OF PERIYAR BASIN USING SWAT
PROJECT REPORT
Submitted by
RITHWIK M (B17CEB45)
AKSHAY MOHAN S (B17CEB05)
SHARAN M (B17CEB52)
GEORGE B THOMAS (B17CEB27)
Department of Civil Engineering
T.K.M. College of Engineering, Kollam- 691005
A. P. J. ABDUL KALAM TECHNOLOGICAL UNIVERSITY
2021

THANGAL KUNJU MUSALIAR COLLEGE OF ENGINEERING
KOLLAM, KERALA
DEPARTMENT OF CIVIL ENGINEERING
CERTIFICATE
This is to certify that this report entitled FLOOD MODELING OF PERIYAR BASIN USING
SWAT is an authentic report of the project done by
RITHWIK M (TKM17CE105)
AKSHAY MOHAN S (TKM17CE013)
SHARAN M (TKM17CE121)
GEORGE B THOMAS (TKM17CE062)
in partial fulfillment of the requirements for the award of the Degree of Bachelor of Technology
in Civil Engineering by the APJ Abdul Kalam Technological University during the year 2021.
Guide: Guide: Co-ordinator: Head of the Department:
Dr. Adarsh S. Dr. Muhammed Siddik A. Dr. Priya K L Dr. Bindu S.
Associate Professor Assistant Professor Assistant Professor Professor
Department of Civil Engg. Department of Civil Engg Department of Civil Engg. Department of Civil Engg.
T.K.M.C.E., Kollam T.K.M.C.E., Kollam T.K.M.C.E., Kollam T.K.M.C.E., Kollam

ACKNOWLEDGEMENT
We consider ourselves privileged to express my gratitude and respect towards all
those who guided me through the completion of the project.
We would like to express our heartfelt gratitude to our Guides, Dr.
Adarsh S, Associate Professor, Department of Civil Engineering and Dr.
Muhammed Siddik A, Assistant Professor, Department of Civil Engineering
for providing encouragement, constant support and guidance which was of
help to complete this report successfully.
We would like to extend our sincere thanks to Dr. Bindu S, Head of
the Department of Civil Engineering and Dr. Sirajuddin M, former HOD,
for their great help in completing this project.
We would also like to express our gratitude to Dr. T. A. Shahul
Hameed, Principal, T.K.M.C.E., for providing us with all the facilities for
carrying out this project.
We are also indebted to all the faculty members of Civil Engineering
Department, T.K.M.C.E, friends and people who took their time to help us.

i
ABSTRACT
The present study is based on SWAT Hydrological model based on calibration and validation
of monthly surface runoff data of Periyar basin at Kerala. The SWAT Model was prepared on
ArcGIS 10.2 software with the help of SWAT Plugin. For the preparation of SWAT Model,
satellite data and geographical information data is used. The study focus on developing a co-
relation between the model performance based on calibration and observer discharge from the
watershed. Land Use/Land Cover, soil map, meteorological data has been procured for the
model preparation. The Digital Elevated Model (DEM) of Periyar Basin, water shed
delineation is done and Hydrological Response Unit (HRUs) are generated. The Number Of
HRUs are 33 and the Number of Subbasins are 21. From the SWAT Simulation, the output
parameters of Periyar Basin, such as Monthly Precipitation and surface runoff were estimated
as 61.202 mm and 56.943 mm respectively and the average curve number was found to be
68.51. The calibration of the SWAT Model is done Using Soil Water Assessment Tool
(SWAT-CUP 2012) which is used for the automatic calibration of Swat simulated model.
Keywords: Rainfall; Runoff; SWATCUP; Calibration; Validation; HRU; Subbasin;

ii
CONTENTS
LIST OF FIGURES ..................................................................................................................... iv
LIST OF TABLES ....................................................................................................................... vi
Chapter 1 ....................................................................................................................................... 1
INTRODUCTION......................................................................................................................... 1
Chapter 2 ....................................................................................................................................... 3
LITERATURE REVIEW............................................................................................................. 3
2.1 SUMMARY.......................................................................................................................... 5
Chapter 3 ....................................................................................................................................... 6
THEORETICAL BACKGROUND............................................................................................. 6
3.1 THEORETICAL BACKGROUND...................................................................................... 6
3.2 SCS CURVE NUMBER FOR RUNOFF ESTIMATION.................................................. 13
3.3 HARGREAVES MODEL FOR ET ANALYSIS ............................................................... 14
Chapter 4 ..................................................................................................................................... 17
STUDY AREA AND DATA....................................................................................................... 17
4.1 STUDY AREA ................................................................................................................... 17
4.2 DEM.................................................................................................................................... 17
4.3 LU/LC ................................................................................................................................. 18
4.4 SOIL MAP .......................................................................................................................... 19
4.5 METEOROLOGICAL DATA............................................................................................ 20
4.6 PRECIPITATION............................................................................................................... 20
4.7 TEMPERATURE................................................................................................................ 20
4.8 DISCHARGE...................................................................................................................... 21
Chapter 5 ..................................................................................................................................... 22

iii
METHODOLOGY...................................................................................................................... 22
5.1 DEM MOSAICKING AND CLIPPING............................................................................. 23
5.2 NEW PROJECT SETUP .................................................................................................... 25
5.3 WATERSHED DELINEATION ........................................................................................ 26
5.4 HRU ANALYSIS AND SWAT SIMULATION................................................................ 27
5.5 CALIBRATION.................................................................................................................. 29
5.5.1 EDITING INPUT PARAMETERS ............................................................................. 33
5.6 VALIDATION.................................................................................................................... 37
Chapter 6 ..................................................................................................................................... 38
RESULT AND DISCUSSION.................................................................................................... 38
6.1 SWAT OUTPUT AND ANALYSIS .................................................................................. 38
6.2 CALIBRATION RESULT AND ANALYSIS................................................................... 39
6.3 VALIDATION RESULT AND ANALYSIS ..................................................................... 42
Chapter 7 ..................................................................................................................................... 44
CONCLUSION............................................................................................................................ 44
Chapter 8 ..................................................................................................................................... 46
REFERENCES............................................................................................................................ 46

iv
Sl
No. LIST OF FIGURES
Page
No:
3.1
3.2
4.1
5.1
5.2
5.3
5.4
5.5
5.6
5.7
5.8
5.9
5.10
A watershed map showing Land use map and
stream network
Representation of Hydrologic cycle
DEM Model
SWAT process flowchart
Layer properties
Mosaic to New Raster
Merged raster output- DEM
ArcToolbox selection
New project setup
Process of watershed delineation and calculation
of subbasin parameters
HRU Analysis
SWAT Simulation
Flowchart of calibration and Validation
7
7
18
22
23
24
24
25
26
27
28
29
30

v
5.11
5.12
5.13
5.14
5.15
5.16
5.17
5.18
6.1
6.2
6.3
6.4
6.5
Selection of New Project
Selection of TxtInOut
Selection of SUFI-2 as project type
Editing parameters and simulation nos
File.Cio edits
Addition of Calibration data from WRIS
SUFI2_extraxt_rch.def edits
Successful simulation run of SWAT-CUP
Inferences from hydrologic cycle
Calibration 95PPU chart
calibration 95PPU chart with higher values at Y-
axis
calibration 95PPU chart peaking at 28th
simulation
Validation 95PPU chart peaking at 73rd sim.
31
32
33
33
34
35
36
36
38
40
41
41
43

vi
Sl
No. LIST OF TABLES
Page
No:
5.1
6.1
6.2
6.3
6.4
Excel sheet representation of WRIS data
Calibration Simulations, no: of iterations and
output values
Comparison between faulty previous simulation
run and accurate run later, after modifications
Validation results
Appendix Table
35
39
42
43
43

1
Chapter 1
INTRODUCTION
The importance and availability of water has become a global concern (Bhattacharyya et al.,
2015). Its indispensable for water resource assessment and proper management. The average
annual precipitation of Kerala is about 3000mm. The South-west and North-east monsoons do
play an important factor in controlling the rainfall of the state. In Kerala, 90% of the rainfall
occurs mainly during six monsoon months. Heavy discharge rate occurs in all rivers in Periyar
basin because of the high intensity storms which occurs commonly during the monsoon period.
Due to the heavy precipitation and steep slopes in topography and undulating terrain, water
flows into the main rivers by numerous streams and water courses (CWC Report of Kerala).
The water balance is important tool to correlate geology, land use, hydrology, climate and soil
with the ground water recharge and storage (Adie et al. 2012). The hydrological process is a
very complicated process and their proper comprehension is essential. There are different
modes of watershed simulations for analysis and time step model have proven their accuracy
(Ercan et at., 2014). Among these, SWAT (Soil Water Assessment Tool) Models have been
used to analyse the effect of climatic conditions, topography and other factors and to identify
the parameters. Model calibration is a process that helps to provide the estimates of model
parameters and to develop a relation with model and prototype. Sensitivity analysis is done in
order to find the model parameters and its used to get more accurate results (Gassman et al.
2007).The sensitivity analysis is of two types (manual v/s automatic) and also about number
of parameters used (one-factor-at-a-time v/s global sensitivity analysis).Both manual and
automatic methods are present for calibration and validation process and also the automatic
calibration and validation of SWAT Models using the automatic methods using numerical
optimisation techniques (like Uncertainty in Sequential Uncertainty Fitting - SUFI2) are
commonly used. For automatic calibration, its capable of implementing multiple objectives

2
employs parameterisation which helps to reduce number of calibration parameters (Bekele and
Nickolow 2007) .For the manual calibration, expert opinion is need for selecting, varying
parameters and also extensive knowledge about the watershed is required (Gassman et al.
2007).
Hydrological modelling and studies of Periyar basin is comparatively low and also Kerala has
witnessed flood during the last few years (2017-2018). The extreme change in weather have
resulted in scarcity of water and also disasters like flood etc. The increasing global warming
and heat have resulted in rise in sea level and has also resulted in unpredictable rainfall and
sudden change climatic conditions. For the proper management, conservation and utilisation
of water sources it’s important to study the rainfall runoff behaviour. Hydrological modelling
of Periyar basin is useful for the water resources engineers, hydrological community,
agricultural management and climate change concerns, and governmental efforts in controlling
extreme natural hazards such as droughts and floods.
1.1 OBJECTIVES
• To create flood models for Periyar basin using ArcSWAT and ArcGIS
• To calibrate the model to an acceptable accuracy and validate it
• To draw conclusion from the final comparative model

3
Chapter 2
LITERATURE REVIEW
Ben Salah et al., (2016) The SWAT model was developed for Wadi Hatab basin,Central Tunisia.
Daily and Monthly water flow and sediment fluxes were simulated for the study region. The Wadi
Hatab Basin was affected with Flash floods and other calamities. The model was calibrated and
validated from 1987-1988 and 1989-1990 respectively. The co coefficient of determination (R2
)
and Nash and Sutcliffe efficiency (NSE) were taken into consideration. R2
ranged from 0.54-0.61
and NSE ranged from 0.52 and 0.61 for calibration and validation respectively. The model was
calibrated on daily basis .The model has satisfactory results on running with monthly basis. The
results prove an agreement with the observed and simulated data. The sediment yield was 1.15 and
5.37 t/ha/year, during the year for calibration and validation respectively.
Himanshu et al., (2016) The study is based on using SWAT (Soil Water Assessment tool) and
GIS in order to create a watershed model of Ken basin of Central India. The model is used to
analyse the sediment yield, hydrology, water balance and to identify the sensitive parameters of
the Ken Basin. On the basic of the Land Use/Land cover map, soil map and slope map the basin
was divided into 10 sub-basins comprising of 143 hydrological response units. The model was
calibrated and validated for monthly and daily basis. R2
value was taken to consideration for
sediment simulation. The calibration of the SWAT Model was done from (1985-1995) and
validated from (1996-2005). The Runoff simulation and sediment simulation provided satisfactory
results (R2
=0.766 and 0.789 for calibration and validation of runoff for daily basis). The model
proved more efficient with monthly data (R2
=0.946 and 0.959).From the analysis of SWAT
Model, evapo-transpiration was found to be more predominant factor for about 44.6% of average
annual precipitation falling over the area. The average annual sediment yield of the basin was
found to be 15.41 t/ha/year. This proves the region has the occurrence of high soil erosion rate.

4
Trudel et al., (2016), The journal analysed the uncertainties in results of Hydrological model
calibration of contrasting complexity. It devises the application of Lumped rainfall runoff model
and SWAT to model watersheds. It discussed about the Evaluation of uncertainty in model using
Parameter Solution Procedure (PARASOL). Using the model for Low flow cases and analysing
the results in different data inputs. Calibration of GR4J and SWAT Model using historic river flow
observation was done over a period of 12 years. Different Regional climatic models were used for
evaluating uncertainty model, to determine variations in climatic projections.
Sahoo et al., (2018) Changes in Land Use and Land Cover (LULC) pattern was studied. It gave
insights to hydrological model requirements and its input parameters like DEM, LULC, soil, slope
and weather data etc. DEM data were used for watershed delineation, where watersheds were
divided into several subbasins. The watershed was divided into several sub-watersheds. The soil
maps were classified into different classes and the slope map was classified into five classes. Then
sub-watersheds were subdivided into a number of hydrologic response units (HRU). Overlay
method was used for HRU analysis in the model. Daily minimum and maximum temperature and
precipitation data were utilized for SWAT simulation.
Shiferaw et al., (2018) SWAT tool was successfully used to simulate the hydrological dynamics
of Ilala watershed and SWAT-CUP was used to calibrate and validate the model. Sensitivity
analysis was done to select the most sensitive parameters for further calibration processes. The
model was calibrated using SWAT-CUP SUFI-2 algorithm and the comparison between the
observed and simulated stream flow were in agreement. The 95% PPU bracketed the average
values of observation by 71% during calibration and 74% during validation. R-factor clocked
between 0.5 and 0.6 during calibration and validation. The simulated and observed hydrographs
for calibration (NSE=0.51, R2
=0.54) and validation (NSE=0.54, R2
=0.63) proved to be in
agreement.
Nasiri et al., (2020) The demand and the availability of the water has become a growing concern.
The increasing demand of the water has gained more importance for Iran. Modelled the watershed
and studied and analysed of available water resource locations and are important for the

5
hydrological community. A semi distributed SWAT Model Was developed for Samalqan
watershed, Iran. The streamflow simulation of the watershed model was done for 13 years. The
model was calibrated and validated by using the SWAT-CUP Programme which helps in
automatic calibration and validation of the Hydrological Model.SUFI-2 algorithm was used.
Sensitivity analysis of the model was done with using 26 SWAT Parameters. The model was
calibrated and validated from 2004 -2012 and 2012-2014 respectively. The co coefficient of
determination (R2
) and Nash and Sutcliffe efficiency (NSE) were taken into consideration. For
calibration, R2
ranged from 0.60–0.80 and NSE ranged from 0.80–0.95. The calibration and
validation of the model were done for monthly basis and the results were satisfactory.
2.1 SUMMARY
The above papers discussed about SWAT analysis of a watershed and calibration and validation
using SWAT-CUP. While all studies used SWAT methodology in general, calibration
methodology was changed according to users’ discretion. From a general perspective, we can see
most of the results utilising SUFI-2 ended up giving good results, with satisfactory R2
values.

6
Chapter 3
THEORETICAL BACKGROUND
3.1 THEORETICAL BACKGROUND
Rather than incorporating regression equations to describe the relationship between input and
output variables, SWAT requires specific information about weather, soil properties, and
topography, vegetation, and land management practices occurring in the watershed. The
physical processes associated with water movement, sediment movement, crop growth,
nutrient cycling, etc. are directly modelled by SWAT using this input data.
Benefits of this approach are:
• watersheds with no monitoring data (e.g. stream gage data) can be modelled
• the relative impact of alternative input data (e.g. changes in management practices,
climate, vegetation, etc.) on water quality or other variables of interest can be quantified
• uses readily available inputs. While SWAT can be used to study more specialized
processes such as bacteria transport, the minimum data required to make a run are
commonly available from government agencies.
• is computationally efficient. Simulation of very large basins or a variety of
management strategies can be performed without excessive investment of time or
money.
• enables users to study long-term impacts. Many of the problems currently
addressed by users involve the gradual build-up of pollutants and the impact on
downstream water bodies. To study these types of problems, results are needed from
runs with output spanning several decades.

7
SWAT allows a number of different physical processes to be simulated in a watershed. For
modelling purposes, a watershed may be partitioned into a number of sub watersheds or sub
basins. The use of sub basins in a simulation is particularly beneficial when different areas of
the watershed are dominated by land uses or soils dissimilar enough in properties to impact
hydrology. By partitioning the watershed into sub basins, we can reference different areas of
the watershed to one another spatially.
Figure 3.1: A watershed map showing Land use
map and stream network
Figure 3.2: Representation of Hydrologic cycle

8
The climate of a watershed provides the moisture and energy inputs that control the water balance
and determine the relative importance of the different components of the hydrologic cycle.
The climatic variables required by SWAT consist of daily precipitation, maximum/minimum air
temperature, solar radiation, wind speed and relative humidity. The model allows values for daily
precipitation, maximum/minimum air temperatures, solar radiation, wind speed and relative
humidity to be input from records of observed data or generated during the simulation.
Weather Generator: Daily values for weather are generated from average monthly values. The
model generates a set of weather data for each sub basin. The values for any one sub basin will be
generated independently and there will be no spatial correlation of generated values between the
different subbasins.
Generated Precipitation: SWAT uses a model to generate daily precipitation for simulations
which do not read in measured data. This precipitation model is also used to fill in missing data in
the measured records. The precipitation generator uses a first-order Markov chain model to define
a day as wet or dry by comparing a random number (0.0-1.0) generated by the model to monthly
wet-dry probabilities input by the user. If the day is classified as wet, the amount of precipitation
is generated from a skewed distribution or a modified exponential distribution.
Sub-Daily Rainfall Patterns: If sub-daily precipitation values are needed, a double exponential
function is used to represent the intensity patterns within a storm. With the double exponential
distribution, rainfall intensity exponentially increases with time to a maximum, or peak, intensity.
Once the peak intensity is reached, the rainfall intensity exponentially decreases with time until
the end of the storm
Generated Air Temperature and Solar Radiation: Maximum and minimum air temperatures
and solar radiation are generated from a normal distribution. A continuity equation is incorporated
into the generator to account for temperature and radiation variations caused by dry vs. rainy
conditions. Maximum air temperature and solar radiation are adjusted downward when simulating
rainy conditions and upwards when simulating dry conditions. The adjustments are made so that

9
the long-term generated values for the average monthly maximum temperature and monthly solar
radiation agree with the input averages.
Generated Wind Speed: A modified exponential equation is used to generate daily mean wind
speed given the mean monthly wind speed.
Generated Relative Humidity: The relative humidity model uses a triangular distribution to
simulate the daily average relative humidity from the monthly average. As with temperature and
radiation, the mean daily relative humidity is adjusted to account for wet- and dry-day effects.
Hydrology
As precipitation descends, it may be intercepted and held in the vegetation canopy or fall to the
soil surface. Water on the soil surface will infiltrate into the soil profile or flow overland as runoff.
Runoff moves relatively quickly toward a stream channel and contributes to short-term stream
response. Infiltrated water may be held in the soil and later evapotranspired or it may slowly make
its way to the surface-water system via underground paths. The potential pathways of water
movement simulated by SWAT in the HRU are illustrated in Figure 0.5.
Canopy Storage: Canopy storage is the water intercepted by vegetative surfaces (the canopy)
where it is held and made available for evaporation. When using the curve number method to
compute surface runoff, canopy storage is taken into account in the surface runoff calculations.
SWAT allows the user to input the maximum amount of water that can be stored in the canopy at
the maximum leaf area index for the land cover. This value and the leaf area index are used by the
model to compute the maximum storage at any time in the growth cycle of the land cover/crop.
When evaporation is computed, water is first removed from canopy storage.
Infiltration: Infiltration refers to the entry of water into a soil profile from the soil surface. As
infiltration continues, the soil becomes increasingly wet, causing the rate of infiltration to decrease
with time until it reaches a steady value. The initial rate of infiltration depends on the moisture
content of the soil prior to the introduction of water at the soil surface. The final rate of infiltration
is equivalent to the saturated hydraulic conductivity of the soil. Because the curve number method

10
used to calculate surface runoff operates on a daily time-step, it is unable to directly model
infiltration. The amount of water entering the soil profile is calculated as the difference between
the amount of rainfall and the amount of surface runoff.
Redistribution: Redistribution refers to the continued movement of water through a soil profile
after input of water (via precipitation or irrigation) has ceased at the soil surface. Redistribution is
caused by differences in water content in the profile. Once the water content throughout the entire
profile is uniform, redistribution will cease. The redistribution component of SWAT uses a storage
routing technique to predict flow through each soil layer in the root zone. Downward flow, or
percolation, occurs when field capacity of a soil layer is exceeded and the layer below is not
saturated. The flow rate is governed by the saturated conductivity of the soil layer. Redistribution
is affected by soil temperature. If the temperature in a particular layer is 0o
C or below, no
redistribution is allowed from that layer.
Evapotranspiration: Evapotranspiration is a collective term for all processes by which water in
the liquid or solid phase at or near the earth's surface becomes atmospheric water vapor.
Evapotranspiration includes evaporation from rivers and lakes, bare soil, and vegetative surfaces;
evaporation from within the leaves of plants (transpiration); and sublimation from ice and snow
surfaces. Potential soil water evaporation is estimated as a function of potential evapotranspiration
and leaf area index (area of plant leaves relative to the area of the HRU). Actual soil water
evaporation is estimated by using exponential functions of soil depth and water content. Plant
transpiration is simulated as a linear function of potential evapotranspiration and leaf area index.
Potential Evapotranspiration: Potential evapotranspiration is the rate at which
evapotranspiration would occur from a large area completely and uniformly covered with growing
vegetation which has access to an unlimited supply of soil water. This rate is assumed to be
unaffected by micro-climatic processes such as advection or heat-storage effects. The model offers
three options for estimating potential evapotranspiration: Hargreaves, Priestley-Taylor and
Penman-Monteith.
Lateral Subsurface Flow: Lateral subsurface flow, or interflow, is streamflow contribution which
originates below the surface but above the zone where rocks are saturated with water. Lateral

11
subsurface flow in the soil profile (0-2m) is calculated simultaneously with redistribution. A
kinematic storage model is used to predict lateral flow in each soil layer. The model accounts for
variation in conductivity, slope and soil water content.
Surface Runoff: Surface runoff, or overland flow, is flow that occurs along a sloping surface.
Using daily or sub daily rainfall amounts, SWAT simulates surface runoff volumes and peak runoff
rates for each HRU.
Surface Runoff Volume is computed using a modification of the SCS curve number method. In
the curve number method, the curve number varies non-linearly with the moisture content of the
soil. The curve number drops as the soil approaches the wilting point and increases to near 100 as
the soil approaches saturation. It requires sub-daily precipitation data and calculates infiltration as
a function of the wetting front matric potential and effective hydraulic conductivity. Water that
does not infiltrate becomes surface runoff. SWAT includes a provision for estimating runoff from
frozen soil where a soil is defined as frozen if the temperature in the first soil layer is less than
0°C. The model increases runoff for frozen soils but still allows significant infiltration when the
frozen soils are dry.
Peak Runoff Rate: Predictions are made with a modification of the rational method. In brief, the
rational method is based on the idea that if a rainfall of intensity i begins instantaneously and
continues indefinitely, the rate of runoff will increase until the time of concentration, tc, when all
of the subbasin is contributing to flow at the outlet. In the modified Rational Formula, the peak
runoff rate is a function of the proportion of daily precipitation that falls during the subbasin tc, the
daily surface runoff volume, and the subbasin time of concentration. The proportion of rainfall
occurring during the subbasin tc is estimated as a function of total daily rainfall using a stochastic
technique. The subbasin time of concentration is estimated using Manning’s Formula considering
both overland and channel flow.
Ponds: Ponds are water storage structures located within a subbasin which intercept surface runoff.
The catchment area of a pond is defined as a fraction of the total area of the subbasin. Ponds are
assumed to be located off the main channel in a subbasin and will never receive water from
upstream subbasins. Pond water storage is a function of pond capacity, daily inflows and outflows,

12
seepage and evaporation. Required inputs are the storage capacity and surface area of the pond
when filled to capacity. Surface area below capacity is estimated as a nonlinear function of storage.
Tributary Channels: Two types of channels are defined within a subbasin: the main channel and
tributary channels. Tributary channels are minor or lower order channels branching off the main
channel within the subbasin. Each tributary channel within a subbasin drains only a portion of the
subbasin and does not receive groundwater contribution to its flow. All flow in the tributary
channels is released and routed through the main channel of the subbasin. SWAT uses the
attributes of tributary channels to determine the time of concentration for the subbasin.
Transmission Losses: Transmission losses are losses of surface flow via leaching through the
streambed. This type of loss occurs in ephemeral or intermittent streams where groundwater
contribution occurs only at certain times of the year, or not at all. SWAT uses Lane’s method
described in Chapter 19 of the SCS Hydrology Handbook (USDA Soil Conservation Service,
1983) to estimate transmission losses. Water losses from the channel are a function of channel
width and length and flow duration. Both runoff volume and peak rate are adjusted when
transmission losses occur in tributary channels.
Return Flow: Return flow, or base flow, is the volume of streamflow originating from
groundwater. SWAT partitions groundwater into two aquifer systems: a shallow, unconfined
aquifer which contributes return flow to streams within the watershed and a deep, confined aquifer
which contributes return flow to streams outside the watershed. Water percolating past the bottom
of the root zone is partitioned into two fractions—each fraction becomes recharge for one of the
aquifers. In addition to return flow, water stored in the shallow aquifer may replenish moisture in
the soil profile in very dry conditions or be directly removed by plant. Water in the shallow or
deep aquifer may be removed by pumping.

13
3.2 SCS CURVE NUMBER FOR RUNOFF ESTIMATION
As its most basic requirement, SWAT needs basic meteorological data like precipitation,
temperature, wind gauge data, Solar radiation, relative humidity (some or all of them, depends
upon the research purpose one needs to undertake), soil types and its properties, land use and land
cover data of the study area and most importantly of all, a DEM to work on. From these input,
SWAT can create models of various physical processes like precipitation, evapo-transpiration,
surface runoff etc. The rainfall runoff model used by SWAT here is the United States Department
of Agriculture (USDA) Soil Conservation Service (SCS) curve number method. It is a method of
estimating rainfall excess from rainfall. For modelling purposes, the river basin is divided into a
number of sub-basins and then divided further into a number of HRUs (Hydrological Response
Units. In a river basin, water balance gives the idea of all physiological processes happening in a
river basin. The land phase of the hydrologic cycle demonstrated in SWAT is centred on the water
balance equation. The law of water balance states that the inflows to any water system or area is
equal to its outflows plus change in storage during a time interval (Nasiri et al. 2020). In hydrology,
a water balance equation can be used to describe the flow of water in and out of a system. It is
given by:
𝑆𝑊𝑓 = 𝑆𝑊𝑖 + ∑(𝑃𝑑𝑎𝑦 − 𝑅𝑠𝑢𝑟𝑓– 𝑄𝑠𝑒𝑒𝑝 − 𝐸𝑎 − 𝐷𝑔𝑤)
𝑡
𝑖=1
where SWf = final water content in soil (mm water); SWi = initial water content in soil on i day
(mm water); Rsurf = surface runoff on i day (mm water); Qseep = water entering the unsaturated
zone of soil on i day (mm); Pday = precipitation on day i (mm water); Dgw = return flow on day
i (mm water); and Ea = amount of evapotranspiration on day i (mm water).
To calculate surface runoff, SCS curve number method was used. We used the Hargreaves
method to estimate potential evapotranspiration in SWAT.
The SCS curve equation is described as:
Eq. 3.1
Eq. 3.2

14
𝑄 =
(𝑃 − 𝐼𝑎)
{(𝑃 − 𝐼𝑎) + 𝑆}
Where, Q= Runoff in mm; P= Rainfall in mm; Ia = initial abstraction; S = potential maximum
retention after runoff begins.
The retention parameter varies spatially due to changes with land surface features such as soils,
land use, slope, and management practices. This parameter can also be affected temporally due to
changes in soil water content. It is mathematically expressed as
𝑆 = (25400
𝐶𝑁
⁄ ) − 254
Where, CN is the curve number corresponding to the day and its value is a function of land use
practice, soil permeability, and soil hydrologic group. If the initial abstraction (Ia) is approximated
as 0.2S, then the equation changes to:
𝑄 =
(𝑃 − 0.2𝑆)2
(𝑃 + 0.8𝑆)
⁄
3.3 HARGREAVES MODEL FOR ET ANALYSIS
After gathering available information, potential evapotranspiration (PET) can be computed by
Penman-Monteith, Priestley-Taylor or Hargreaves method. The information required for the
Penman-Monteith method are solar radiation, air temperature, wind speed, and relative humidity
(RH), along with precipitation data. In Priestley-Taylor method, radiation information, air
temperature, and relative humidity are needed, in addition to precipitation data. But in Hargreaves
method only air temperature data is needed in addition. Since this is the easiest PET method and
needs only minimal data compared to the other methods, Hargreaves method was used. Hargreaves
model is one of the most used in research purposes, and is the simplest one for practical use of
analysis of Evapotranspiration. The Hargreaves model is expressed as follows:
Eq. 3.3
Eq. 3.4

15
𝐸𝑇0 = 0.0135 (𝑇 + 17.78)𝑅𝑠
Where ET = potential daily evapotranspiration, mm/day; T = mean temperature, °C; and RS =
incident solar radiation converted to depth of water, mm/day.
There are many methods to measure and evaluate the accuracy of results produced by the model.
The calibration and the validation were carried out using the three commonly statistic objective
functions of a possible nine, namely the coefficient of determination (R2
) and Nash–Sutcliffe
efficiency index (NSE) or Percent bias (PBIAS). The coefficient NSE (efficiency ratio) specifies
to what range the simulated values approximate the observed values. It varies in values ranging
from −∞ to 1. The model effectively replicates most accurately if NSE is close to 1. Determination
coefficient (R2
) is a value between 0 and 1; it is optimal for a value equal to 1, which indicates that
estimated values correspond to the measured actual values. The PBIAS optimal value is zero.
Positive values indicate a pattern of underestimation bias, and negative values indicate an
overestimation bias model (Ben Salah and Abida, 2016). Coefficient of determination R2
is given
by:
𝑅2
=
[∑ (𝑂𝑖 − 𝑂
̅)(𝑃𝑖 − 𝑃
̅)]
𝑛
𝑖=1
∑ [
𝑛
𝑖=1 (𝑂𝑖 − 𝑂
̅) ∑ (𝑃𝑖 − 𝑃
̅)2]
𝑛
𝑖=1
And the Nash–Sutcliffe efficiency index (NSE) is given by:
Eq. 3.5
Eq. 3.6
Eq. 3.7

16
𝑁𝑆𝐸 =
∑ (𝑃𝑖 − 𝑂𝑖)2
𝑛
𝑖=1
∑ (𝑂𝑖 − 𝑂
̅)2
𝑛
𝑖=1
Where Pi is the ith
observation (stream flow), Oi is the ith
simulated value, O is the mean of
observed data, and n is the total number of observations (Nasiri et al. 2020)

17
Chapter 4
STUDY AREA AND DATA
4.1 STUDY AREA
Periyar River is a 244 km long, west flowing river, originating from Western Ghats and draining
to Arabian Sea in the Eranakulam district of Kerala. Its largest tributaries are the Muthirapuzha
River, the Mullayar River, the Cheruthoni River, the Perinjankutti River and the Edamala River.
Its basin size is 5398 km2. As our study area, we’ve selected a portion of the basin within the
geographical coordinates (10.3441, 76.3565), (9.2431, 76.7834), (10.3823, 77.4023), (9.2270,
77.6096). It consists of the whole of Idukki district and the eastern portion of Eranakulam
district, till Neeleswaram point where the outflow is calculated. The entire basin is divided into
34 HR Units and 21 basins, ranging from eastern part of Eranakulam district to Southern part of
Idukki District.
4.2 DEM
Digital Elevation Model is the 3-D rendering of a terrain portrayed by the help of computer
graphics. It provides us with the spatial database of elevation. DEMs are the most common basis
for digitally produced relief maps and are frequently used in geographic information systems.
DEM is represented as a raster data, where it is in the form of a grid of squares. We have acquired
the DEM data for the basin from United State Geographical Survey (USGS),
(https://www.usgs.gov/).
The Periyar basin measures to a vast area of about 5398 square kilometers. The area of DEM
considered for this study is approximately 3949.57 square kilometers. The specific areas are
mentioned in detail under the heading study area. The database requirement for SWAT model

18
includes DEM, soil type, land use, weather (temperature and precipitation) and river discharge
data to establish the water balance. Usually the weather data contains relative humidity and wind
speed but these variants are not considered here due to lack of complete data.
With a proper DEM we can delineate the watershed to find out the networks of river streams, sub-
basins, and other parameters like slopes for HRUs. It helps in understanding the flow behaviour
and flow pattern along the entirety of the area showing the main discharge points. The Periyar
river basin has been divided into 21sub-basins and 33 HRUs based on uniform soil, land use and
slope. The thresholds provided for obtaining the multiple HRUs are Land Use/Soil/Slope were 35
% / 10 % / 35 % respectively.The rainfall and stream discharge data for the period 2001-2010 are
used for the study.
Fig 4.1 DEM Model
4.3 LU/LC
The land use/land cover (LULC) dataset is used to understand the hydrological processes and

19
Governing system. Crop specific digital layers for the preparation of LULC map have been
obtained from the Global Land Cover Facility (GLCF) (https://geog.umd.edu/feature/global-land-
cover-facility-glcf)
The Periyar basin consists of a variety of vegetation from wet and semi evergreen to moist and dry
deciduous areas. Around 35% of the area is forest covered of which some are now utilized for
future development activities. The main activities pertaining the highland areas are plantations,
hydroelectric projects i.e. the Idukki hydroelectric project. The irrigations projects in the area are
centered on the midland portion having paddy fields, coconuts and plantains. The waste lands such
as coastal saline belts and high peaks are only a mere 5-8 % of the entire basin area. It is in the
lower plains that major settlements and industries are situated i.e. the urban area. The region also
has grasslands, forests, plantations and unclassified areas.
According to Venkitaraman V et al. (2014) the areas of settlement according to human activity are
further classified as village, town, commercial and industrial. And the forest region are categorized
into dense forests, dense scrub, open scrub, water bodies. The water bodies present in the
watershed region are either rivers or streams.
4.4 SOIL MAP
Kerala usually has higher annual average rainfall as it is present in the windward side of the
Western Ghats and has significantly high Indian summer monsoon rainfall (ISMR) i.e., from June
to September . Even among the districts of Kerala there are huge variations in maximum and
minimum rainfall. For example the Idukki district located within the Periyar basin received
maximum rainfall of 3555mm which is easily over 100% of the n (Sudheer KP et al. 2019), stated
that the Periyar river basin has mainly three varieties of rock formations such as crystalline rocks,
tertiary and quaternary formations. The sedimentary rock formations predominant in this basin
area are the laterite and alluvium crystalline formed by the stream or river coasts. We required the
Soil data required for our analysis from Food and Agriculture Organisation, FAO
(http://www.fao.org/home/en/ )

20
4.5 METEOROLOGICAL DATA
The meteorological data we used to assess and predict the outflow in the study was hydrological
(discharge) and weather data (temperature, precipitation, relative humidity, wind speed) of the
Periyar basin. We acquired the study area’s meteorological data from the Indian Meteorological
Department, (https://mausam.imd.gov.in/).
4.6 PRECIPITATION
Normal rainfall of that period for other areas which recorded 1852mm. But there are cases reported
where during the first 3 week periods of August in 2018, the recorded rainfall went above 164%
than the normal which is an Extreme Rainfall Event (EREs). The two places that fell victim to this
event was the Peerumedu region which noted above 800mm and the Idukki region with above 700
mm rainfall within 2 days, (Sudheer KP et al., 2019). These kinds of uncertainties can be
determined prior to the event with fair accuracy, provided that we have the necessary rainfall data
of the region covering a few decades time period.
4.7 TEMPERATURE
In the upstream regions like Idukki, situated at a higher location the maximum temperature range
was noted between 25°C to 32°C and the minimum temperature comes between 14°C to 19°C.
Whereas in the lower areas i.e., the downstream region, the maximum temperature recorded was
from 28°C to 32°C and the minimum temperature values varies from 23°C to 26°C, (Sudheer KP
et al., 2019). The temperature data for the time period 1990 to 2013 was obtained.

21
4.8 DISCHARGE
The daily discharge data was acquired through the help of Central Water Commission for a period
of 1979 to 2013. The maximum daily discharge was found to be during the month of august, which
lies between the months of June and September where the average rainfall is much higher. From
this it is evident that we have higher runoff during these periods. It is absolutely necessary for us
to find out the discharge variations in the study area, as the calibration and validation part solely
relies on the even distribution of wet and dry periods of the basin.

22
Chapter 5
METHODOLOGY
Figure 5.1: Flowchart of methodology
The methodology to be followed in the project is as listed below:
A Digital Elevated Model is (DEM) is a specialized database that represents the relief of a surface
between points of known elevation. By interpolating known elevation data from sources such as
ground surveys and photogrammetric data capture, a rectangular digital elevation model grid can
be created. The DEM of our choice was of our study area, Periyar river basin within the
geographical coordinates (10.3441, 76.3565), (9.2431, 76.7834), (10.3823, 77.4023), (9.2270,
77.6096). It consists of the whole of Idukki district and the eastern portion of Eranakulam district,
till Neeleshwaram point where the outflow is calculated.

23
5.1 DEM MOSAICKING AND CLIPPING
The Mosaic tool is used to mosaic multiple input rasters into an existing raster dataset. The existing
raster dataset can be empty or it can contain data. The tool is used to merge rasters that are adjacent
and have the same cell resolution and coordinate system.
1. Determine the number of bands and pixel type of the raster files. (Right-click Table of
Contents, click Properties and the Source tab.) The inputs must have the same number of
bands and same bit depth.
Figure 5.2 Layer properties
2. Open the Mosaic-To-New Raster tool by navigating to Arc Toolbox > Data Management
Tools > Raster > Raster Dataset.
a. Insert the raster files.
Select the output location.
b. Specify a name and extension for the output.

24
c. Specify the pixel type.
d. Specify the number of bands.
Figure 5.3 Mosaic to New Raster
3. Run the tool.
The following image shows the output of a merged raster:
Figure 5.4: Merged raster output-
DEM

25
5.2 NEW PROJECT SETUP
Aftrer creating a new blank document, we chose ArcSWAT option from ArcToolbox.
Figure 5.5: ArcToolbox selection
The SWAT toolbox is displayed then. After setting up the project path, a name was given for
the personal geo-database (a form of access database) under a user-specified project folder. It
is stored in .mdb (geo-database file) format. Geodatabases are relational databases that can also
store geographic features along with normal features. That is, a geodatabase is a collection of
tables whose fields can store a geographic shape (i.e., a point, a line, or a polygon), a string, or
a number and that are related to each other through key fields. Regardless of the number of
tables and relationships in a geodatabase, it is stored in a single file, and its contents can be
explored using database management systems (DBMS). Microsoft Access can extract the
database, and acted as our database management system.

26
Figure 5.6 New project setup
5.3 WATERSHED DELINEATION
Watershed delineation is a common method of locating and delineating the boundaries of
watersheds is by using topographic maps following the basic principle that water runs
downhill. In watershed delineation sub process of SWAT, the first step was to select the DEM.
After DEM input, DEM projection was applied. It is at this point the DEM gets integrated to
SWAT software and is readied for further processing. Since the DEM was of high resolution,
no masking or stream network burning effects were added to enhance the DEM.
In Stream Definition tab, a DEM-based flow direction and accumulation approach was used.
In this step, the built-in ArcHydro tools in ArcGIS processed the network and basin area was
calculated. Calculated basin area was divided into 82578 cells within Automatic Delineation
tool in order to process latter steps in a fast manner by making use of the system’s parallel
processing capabilities. In the next step, the delineator created streams and outlets out of the
DEM.

27
In Outlet and Inlet definition, sub basin outlet was selected. Since Periyar is a west flowing
river, the outlet was selected close to Neeleswaram in Eranakulam district. After selecting that
outlet point once again and inserting it into watershed outlet, the model is finally delineated.
After delineation, by applying the ‘calculate subbasin parameters’ function, we finally got the
no: of sub-basins as 21 and the number of HR Units as 34. This marked the end of Automatic
Watershed Delineation process.
Figure 5.7: Process of watershed delineation and calculation of subbasin parameters
5.4 HRU ANALYSIS AND SWAT SIMULATION
In HRU Analysis, the already delineated DEM was subjected to other relevant data inputs, such
as Land Use and Land Cover (LULC) and Soil Data. LULC data was obtained from SWAT

28
Global Database, which was inserted via raster input option. Once it was inserted, it was
processed further by the in-plugin process.
Under LULC tab, Land Use and Land cover data of the year 2010 was selected. After it was
inserted and was classified based on land use format with different colours, we entered the land
use data manually. Similarly under Soil data tab, soil raster was inserted, selected SNUM type
and soil data classification was inserted manually. A slope class type of 5 nos were selected.
Upon applying HRU overlay, thses changes were collectively acted, and the processing took a
good 30 minutes, before it was finally ready to go into weather data input.
Figure 5.8: HRU Analysis
After careful input of weather data (precipitation, temperature, wind speed, solar radiation and
relative humidity), the SWAT process files were updated. Necessary changes were done where
we selected desired the PET method to be Hargreaves method. After this, we moved on to the
simulation window, where the simulation was run successfully using the hydrologic cycle
parameters are generated by SWAT simulation.

29
Figure 5.9: SWAT Simulation
5.5 CALIBRATION
SWAT input parameters are process based and must be held within a realistic uncertainty
range. The first step in the calibration and validation process in SWAT is the determination of
the most sensitive parameters for a given watershed or sub watershed. We determined which
variables to adjust based on inferences or on sensitivity analysis. Sensitivity analysis is the
process of determining the rate of change in model output with respect to changes in model
inputs (parameters). It is necessary to identify key parameters and the parameter precision
required for calibration. In a practical sense, this first step helps determine the predominant
processes for the component of interest. Two types of sensitivity analysis are generally

30
performed: local, by changing values one at a time, and global, by allowing all parameter
values to change. The two analyses, however, may yield different results. Sensitivity of one
parameter often depends on the value of other related parameters; hence, the problem with one-
at-a-time analysis is that the correct values of other parameters that are fixed are never known.
The disadvantage of the global sensitivity analysis is that it needs a large number of
simulations. Both procedures, however, provide insight into the sensitivity of the parameters
and are necessary steps in model calibration.
Figure 5.10: Flowchart of calibration and Validation
(Source: SWAT-CUP manual, Abbaspour)
Calibration is an effort to better parameterize a model to a given set of local conditions, thereby
reducing the prediction uncertainty. Model calibration is performed by carefully selecting
values for model input parameters by comparing model predictions (output) for a given set of
assumed conditions with observed data for the same conditions.
In SWAT-CUP, a new project setup option was selected. After this, we selected the right
ArcGIS version and SWAT-CUP version, which are necessary for enhancing smooth parallel
process capabilities.

31
Figure 5.11: Selection of New Project
Selection of TxtInOut file was done later. At this stage, we had to select the SWAT project
file. TxtInOut is a file which is located in the SWAT project directory. SWAT output database
is stored in .mxd format as well as an Access database. It contains the SWAT process data,
result and characteristics, which acts as the input for calibration and validation process in
SWAT-CUP.

32
Figure 5.12: Selection of TxtInOut
Here, we selected SUFI-2 (Sequential Uncertainty Fitting) calibration method from the list of
uncertainty methods. The SUFI-2 algorithm in the SWAT-CUP software package was used
for model calibration, validation, sensitivity, and uncertainty analysis. In SUFI-2, parameter
uncertainty accounts for all sources of uncertainties such as uncertainty in driving variables
(e.g., rainfall), conceptual model, parameters, and measured data. The degree to which all
uncertainties are accounted for is quantified by a measure referred to as the P-factor, which is
the percentage of measured data bracketed by the 95% prediction uncertainty (95PPU). It is
complimented by R-factor, another measure quantifying the strength of a
calibration/uncertainty analysis, which is the average thickness of the 95PPU band divided by
the standard deviation of the measured data. SUFI-2 hence brackets most of the measured data
with the smallest possible uncertainty band. The combination of P-factor and R-factor together
indicate the strength of the model calibration and uncertainty assessment, as these are
intimately linked.

33
Figure 5.13: Selection of SUFI-2 as project type
5.5.1 EDITING INPUT PARAMETERS
1. No: of parameters were selected as 4 (default) and no: of simulations were selected as 100.
Figure 5.14: Editing parameters and simulation numbers

34
2. Edited the Number of years simulated to be 25 (taking calibration period of 25 years, as per
the WRIS data of Neeleswaram from 1990-2014). Beginning Julian day as 1 and Ending Julian
day as 365 (implying the number of days).
Figure 5.15: File.Cio edits
3. Inserted the yearly river discharge obtained from WRIS for Neeleswaram CWC point
Observed_rch.txt

35
Figure 5.16: Addition of Calibration data from WRIS
Following is the excel table of WRIS data used for calibration and validation. This is the only
external data input at this stage, and it is the flow data of Periyar river at Neeleswaram CWC
point, in Eranakulam district. It is represented in cumecs units.
Table 5.1: Excel sheet representation of WRIS data

36
4. In SUFI2_extract_rch.def file, defined how parameters should be extracted from output.rch
file. Selected the first reach to be 1 and the reach of variable column number as 7 (it is the
reach/sub basin which occupies outlet point)
Figure 5.17: SUFI2_extraxt_rch.def edits
5. After selecting the objective function of choice as R2
(Coefficient of co-relation), we proceed
to the calibration process. The windows closed, and a cmd prompt window opened, indicating
calibration process.
Figure 5.18: Successful simulation run of SWAT-CUP

37
5.6 VALIDATION
Calibration and validation are typically performed by splitting the available observed data into
two datasets: one for calibration, and another for validation. Data are most frequently split by
time periods, carefully ensuring that the climate data used for both calibration and validation
are not substantially different, i.e., wet, moderate, and dry years occur in both periods.
For our validation, the steps were same as that of calibration, with the only exception being
dividing the years into two: one before the peak calibrated year and the other after the peak
calibrated year. We selected this year to be 2000-2001. Barring the NYSKIP of 2 years, hence
the calibration period was 1992-2000 and the validation period was 2001-2013. Hence, with
this modified data, the validation process was completed.

38
Chapter 6
RESULT AND DISCUSSION
6.1 SWAT OUTPUT AND ANALYSIS
The hydrological cycle obtained as a result of SWAT process are as follows:
Figure 6.1: Inferences from hydrologic cycle
From the system produced SWAT plugin-simulation results, it was observed that:
• Precipitation may be too high (> 3400 mm)
• Surface runoff to precipitation is at 23%, which is satisfactory.
• Groundwater ratio may be low
• Lateral flow is greater than groundwater flow, may indicate a problem
• Water yield may be excessive

39
6.2 CALIBRATION RESULT AND ANALYSIS
Ideal P-factor value is destined to be 1, indicating 100% bracketing of the measured data,
hence capturing or accounting for all the correct processes. If the value is close to 1, it
indicates the bracketing is close to perfection. For us, the P-factor obtained was 0.77.
This indicates a bracketing success percentage of more than three-quarters, i.e., 77%.
Also, value of R-factor should ideally be near zero, hence coinciding with the measured
data. It indicates a perfect quantifiable strength of a calibration/uncertainty analysis,
which is the average thickness of the 95PPU band divided by the standard deviation of
the measured data. Our R- factor obtained was 0.37 indicates a good quantifiable
strength of a calibration/uncertainty analysis, which is the average thickness of the
95PPU band divided by the standard deviation of the measured data.
The coefficient of determination (r2
) is a statistical measure of the strength of the
relationship between the relative movements of two variables. The values of ‘r’ range
between -1.0 and 1.0. Ideal r2
value of 1 indicates perfect correlation. We got a r2 value
of 0.98, which is very much close to 1, and shows the data are in near-perfect co-relation.
Table 6.1: Calibration Simulations, number of iterations and output values
Sl. No P-factor R-factor r2
No: of iterations
1 0.45 0.51 0.98 250
2 0.59 0.32 0.98 200
3 0.67 0.23 0.98 400
4 0.61 0.40 0.99 450
5 0.58 0.14 0.98 400
6 0.66 0.10 0.98 750
7 0.73 0.08 0.99 1000

40
8 0.71 0.11 0.97 1000
9 0.69 0.19 0.97 800
10 0.69 0.22 0.98 1000
11 0.77 0.37 0.98 1000
12 0.70 0.16 1.0 750
13 0.57 0.29 0.96 800
14 0.52 0.27 0.97 650
This table shows the most acceptable values of P-factor, R-factor and R2
was
encountered for simulation no: 11.
Figures given below (5.2-5.5) are the final outputs of calibration. 95PPU is a chart
plotted between flow in cumecs and no: of simulations for 7th
reach/subbasin, where the
outlet point is situated
Figure 6.2: Calibration 95PPU chart
X- Monthly time period in numbers
Y-
Discharge
data

41
Figure 6.3: Calibration 95PPU chart
Figure 6.4: Calibration 95PPU chart showing to at 28th simulation
Y-
Discharge
data
Y-
Discharge
data

42
Table 6.2: Comparison between faulty previous simulation run and accurate run
later, after modifications
NS Best simulation
No:
No: of
iterations
Previous Run - 0.19 0.27 0.50 - - 50
Current Run 11 0.77 0.37 0.98 0.94 411 1000
Errors in previous run was accounted due to the excess of climatic data points in
precipitation data, which was rectified later. After rectification, the run yielded better
result with R2
value 0.98.
6.3 VALIDATION RESULT AND ANALYSIS
For us, the P-factor obtained was 0.92. This indicates a satisfactory value. Our R-
factor obtained was 0.27, indicates a good quantifiable strength of a
calibration/uncertainty analysis, which is the average thickness of the 95PPU band
divided by the standard deviation of the measured data.
Observed discharge data has successfully fallen in the range of 95PPU plot with a
coefficient of determination (r2
) of 0.61, which is satisfactory.

43
Figure 6.5: Validation 95PPU chart peaking at 73rd simulation
Table 6.3: Validation results
NSE No: of
iterations
1 0.92 0.27 0.61 0.54 250
Validation result for the period 2001-2013 is shown in the table above.
Table 6.4: Appendix Table
Sl.
No
Objective
Function
Definition Range
1 P-Factor The P-factor represents the fraction of the measured
data bracketed by the 95PPU band
0 to 1
2 R-Factor The ratio of the average width of the 95PPU band
and standard deviation of the measured variable
0 to 1
3 R2
Coefficient of determination 0 to 1
4 NSE Nash-Sutcliffe Efficiency - ∞ to 1
Y-
Discharge
data

44
Chapter 7
CONCLUSION
Hydrological modelling of Periyar river basin was developed using SWAT, for the
simulation of Discharge. The model was simulated and surface runoff to
precipitation is 23% and Average Curve Number was found to be 68.51. The model
performance was analysed by the calibration and validation process using
SWATCUP. The coefficient of determination (R2
) was taken as the main objective.
The p-factor obtained was 0.77, R-factor obtained was .37 and r2
value of 0.98. For
validation, the p-factor obtained was 0.92, R-factor obtained was 0.27 and r2
value
of 0.61 .Number of simulation of both the calibration and validation are 1000 and
the 95PPU Plot was obtained for sub-basin no 7. The results of calibration and
validation clearly proves a relation between the observed and simulated flow rates.
The Results of the model simulation were satisfactory and the simulated discharge
was similar with observed discharge rates. The future prediction of climate data will
help for the discharge analysis and would help for the prediction of runoff. Our study
results indicate that the SWAT and SWAT-CUP models are useful in forecasting
flow and performing uncertainty and sensitivity analyses. This proves the reliability
and the efficiency of the SWAT Model for Periyar basin, Kerala. The model can be
used to study the various effects and can be used for the planning and management
of hydrological sources and for disaster management and mitigation activities. It is
evident that the rainfall pattern has become unpredictable over time. This study
model can be used in further assessment of climate change and land use/land cover
impact assessment on river basin.

45
Change in built–up area, deforestation, unchecked urbanization and land use changes
are factors contributing towards global climate change, and we believe that this
unpredictability might have caused a lapse in anticipating frequent flash floods of
the years 2018 and 2019.

46
Chapter 8
REFERENCES
1. Adie, D. B., Ismail, A., Muhammad, M. M., & Aliyu, U. B. (2012). “Analysis of
the water resources potential and useful life of the Shiroro Dam, Nigeria”.
Nigerian Journal of Basic and Applied Sciences, 20(4), pp. 341-348.
2. Bekele, E. G., & Nicklow, J. W. (2007). “Multi-objective automatic calibration
of SWAT using NSGA-II”. Journal of Hydrology, 341(3-4), pp. 165-176.
3. Ben Salah, N.C., Abida, H. “Runoff and sediment yield modeling using SWAT
model: case of Wadi Hatab basin, central Tunisia”. Arab J Geosci 9, 579 (2016).
https://doi.org/10.1007/s12517-016-2607-3
4. Bhattacharyya, D., Clarke, J. T., Bertaux, J. L., Chaufray, J. Y., & Mayyasi, M.
(2015). “A strong seasonal dependence in the Martian hydrogen exosphere”.
Geophysical Research Letters, 42(20), pp. 8678-8685.
5. Ercan, M. B., Goodall, J. L., Castronova, A. M., Humphrey, M., & Beekwilder,
N. (2014). “Calibration of SWAT models using the cloud”. Environmental
Modelling & Software, 62, pp. 188-196.
6. Gassman, P. W., Reyes, M. R., Green, C. H., & Arnold, J. G. (2007). “The soil
and water assessment tool: historical development, applications, and future
research directions”. Transactions of the ASABE, 50(4), pp. 1211-1250
7. Himanshu, S. K., Pandey, A., & Shrestha, P. (2017). “Application of SWAT in
an Indian river basin for modeling runoff, sediment and water balance”.
Environmental Earth Sciences, 76(1), pp. 1-18.
8. Mateeul H, Memon A, Sher M, Siddiqi P, Jillani R, (2012) “Techniques of
Remote Sensing and GIS for flood monitoring and damage assessment: A case
study of Sindh province, Pakistan”, The Egyptian Journal of Remote Sensing and
Space Science, Volume 15, Issue 2, pp 135-141
9. Nasiri, S., Ansari, H., & Ziaei, A. N. (2020).”Simulation of water balance
equation components using SWAT model in Samalqan Watershed (Iran)”.
Arabian Journal of Geosciences, 13, pp. 1-15

47
10. Sahoo S, Dhar A , Debsarkar A and Kar A, (2018), “Impact of water demand on
hydrological regime under climate and LULC change scenarios”, Environmental
earth science (2018) 77:341
11. Shiferaw H, Gebremedhin A, Gebretsadkan T and Zenebe A, (2018), “Modeling
hydrological response under climate change scenarios using SWAT model: the
case of Ilala watershed, North Ethiopia”, Modeling Earth systems and
Environment 4: pp.437-449.
12. Sudheer K, Bhallamudi M, Narasimhan B, Thomas J, Bindhu V.M, Vema V K,
Kurian C (2019). “Role of dams on the floods of August 2018 in Periyar River
Basin, Kerala”, Current Science 116, pp. 780-794. 10.18520/cs/v116/i5/780-794.
13. V. Venkatraman, Er. P. Selvan , Dr. S. Chandran (2014), “Land Use and Land
Cover Change Detection of Periyar Main Canal Command through Remote
Sensing Using Multi-Temporal Satellite Data”, International Journal Of
Engineering Research & Technology (IJERT) Volume 03, Issue 06 (June 2014)
14. Trudel M, Doucet-Genereux P. Land Leconte B (2016), “Assesing River Low-
Flow Uncertainties Related To Hydrological Model Calibration and Structure
Under Climate Change Conditions”, Climate 2017, pp. 5- 19

Flood modelling of Periyar basin using SWAT

Recommended

Recommended

More Related Content

What's hot

What's hot (20)

Similar to Flood modelling of Periyar basin using SWAT

Similar to Flood modelling of Periyar basin using SWAT (20)

Recently uploaded

Recently uploaded (20)

Flood modelling of Periyar basin using SWAT