A13

EFFECTS OF LULC CHANGE ON THE HYDROLOGIC PROCESSES OF

MANABA WATERSHED AND RIVER MOUTH: A SWAT MODEL ANALYSIS

JANN MERR C. JAMINAL

JAMLECH NIÑO JOSHUA G. JIMENEZ
NELVE Q. TAER

A Research Proposal Presented to

Bianca Baldassarri, Ph.D.
Central Visayan Institute Foundation

In Partial Fulfillment of the Requirements

for the Science, Technology, Engineering and
Mathematics (STEM) Research Project

APRIL 2024
 TABLE OF CONTENTS

PAGE
Table of Contents ……………………..………………………………….……………….…………… ii
List of Figures …………………………………………………………………………………………….. iii
List of Tables ……………………………………………………………………………………………… iv
List of Abbreviations ………….……………………………..……………………….………………. v
CHAPTER 1. INTRODUCTION ……………………………………………………………………. 1
1.1. Background and Rationale.……….………………………….............................. 1
1.2. Objectives of the Study………………………………………............................... 2
1.3. Significance of the Study ……………………………………............................... 2
1.4. Scope and Limitations of the Study………………………………………………….. 3
CHAPTER 2. REVIEW OF RELATED LITERATURE ………………………………………… 4
2.1. LULC Change.……….………………………….................................................... 4
2.2. SWAT……………………………………………………………………………………………….. 4
2.3. LULC……………………………………................................................................ 5
2.4. Methods and Models……………………………………………………………………….. 5
2.5 Watershed………………………………………………………………………………………… 7
2.6 SWAT Applications and LULC Change Analysis…………………………………… 9
CHAPTER 3. MATERIALS AND METHODS …………………………………………………… 11
3.1 Materials …………………....…………………………………................................... 11
3.2 Methods…………………….………………………………........................................ 17
LITERATURE CITED.......…………………………………………………………………………….…. 31

ii
 LIST OF FIGURES

FIGURE TITLE PAGE

3.1 Watershed map of Garcia Hernandez………………………………………. 13

3.2 Coordinated map of Manaba River………………………………………….. 14

3.3 Length map of and width of Manaba River………………………………. 15

3.4 SWAT model flowchart……………………………………………………………. 17

3.5 DEM of Bohol region around Manaba watershed……………………. 19

3.6 Slope map of Manaba watershed generated from the DEM……. 20

3.7 Digital soil map of the world……………………………………………………. 21

3.8 Weather station location in the study domain…………………………. 24

3.9 River gauge location for model calibration and validation……….. 25

iii
 LIST OF TABLES

TABLE TITLE PAGE

3.1 Model inputs and data sources for the SWAT simulations………. 16

3.2 User-defined slope classes configured in the SWAT set-up……… 20

3.3 Look-up table used in SWAT on the ESA Sentinel-2 LULC maps.. 22

3.4 Pearson correlation table………………………………………………………… 27

3.5 Output parameters………………………………………………………………….. 28

iv
 LIST OF ABBREVIATIONS

ALOS Advanced Land Observing Satellite

BSWM Bureau of Soils and Water Management

CLUP Comprehensive Land Use Plan

DEM Digital Elevation Model

DENR Department of Environment and Natural Resources

DPWH Department of Public Works and Highways

DSMW Digital Soil Map of the World

ESA European Space Agency

FAO Food and Agriculture Association

FP-IT Forward Processing-Instrument Teams

GEWEX Global Energy and Water Exchange Project

GMAO Global Modeling and Assimilation Office

HRUs Hydrologic Response Units

JAXA Japan Aerospace Exploration Agency

LULC Land Use Land Cover

Modern Era Retrospective-Analysis for Research and

MERRA
Applications

MPDC Municipal Planning and Development Coordinator

MUSLE Modified Universal Soil Loss Equation

NASA National Aeronautics and Space Administration

NSE Nash-Sutcliffe model 2 Efficiency

PBIAS Percent Bias

POWER Prediction Of Worldwide Energy Resources

v
 Panchromatic Remote-sensing Instrument for Stereo
PRISM
Mapping

SRB Surface Radiation Budget

SWAT Soil and Water Assessment Tool

USLE Universal Soil Loss Equation

VAR Vector Auto Regression

vi
 CHAPTER 1

INTRODUCTION

1.1 Background and Rationale

Land Use Land Cover (LULC) integrates socio-economic and physical aspects of
land, such as agriculture and vegetation, to understand human activities and natural
elements on landscapes [1]. Geologic processes, such as erosion and weathering, and
varying weather patterns, are some of the natural factors. Changing demands in the
economy also influence LULC changes. These LULC changes have a significant impact
on the watershed processes, particularly in the hydrological system [2, 3, 4]. Although
field observations are crucial, they are often costly and sparse, both spatially and
temporally. Numerical models, like the Soil and Water Assessment Tool (SWAT),
provide a cost-effective means to interpolate observations and forecast
environmental changes, proving essential for watershed management [5, 6, 7].

SWAT is particularly effective in assessing the effects of land use, soil erosion,
and pollution on watersheds. In the Manaba Watershed in Bohol, Philippines, LULC
changes could significantly impact the water quality and quantity of the Manaba River,
the third cleanest in the country, affecting ecosystem services and local livelihoods [8,
9].

This study employs SWAT to assess the impact of LULC changes in the Manaba
Watershed on the river's discharge, including conducting a correlation analysis
between rainfall and discharge. What are the effects of LULC change on the Hydrologic
processes of Manaba Watershed and River mouth? What is the relationship between
rainfall and river discharge in the Manaba Watershed? The findings aim to provide
insights into the effects of human-induced changes and guide water resource
management strategies for the benefit of the river's ecosystems and surrounding
communities.

1
1.2 Objectives of the Study

The main objective of this research is to investigate the effects of LULC change
in Manaba Watershed on the river discharge of the Manaba River, using the SWAT.
Specifically, it aims to:

● to determine the changes in hydrologic parameters (i.e. sediment

yield in t/ha, river discharge in m3/s, suspended load in metric tons,
and suspended sediment concentration in mg/L) of the Manaba
Watershed when running two separate simulations using LULC 2017
and LULC 2022;
● to investigate whether these changes are statistically significant;
● to conduct a correlation analysis between rainfall and river discharge
to understand the influence of climatic factors on river flow;
● to provide science-based recommendations for the management of
Manaba Watershed.

1.3 Significance of the Study

This study aims to reveal how changes in LULC affect the hydrological
processes of the Manaba Watershed and river mouth. The findings will provide
important information that can be used to improve land management practices,
formulate better policies, and develop strategies for reducing the risk of disasters in
the area. Understanding the impact of LULC changes on river discharge can help local
communities and farmers in Garcia-Hernandez make more informed decisions about
land use practices, water use, and agriculture. Specifically, the ones assigned in
formulating the Comprehensive Land Use Plan (CLUP) (“Reclassification of Lands”,
1991) can gain insights from the results of this study.

The findings of this research can inform water resource management

strategies and policy decisions in the Manaba Watershed and River, helping to ensure
sustainable water use and mitigate the impacts of LULC changes on water availability.
The study can contribute to efforts to conserve the integrity of the Manaba River as
the third cleanest river in the Philippines and restore natural habitats and ecosystems
2
in the Manaba Watershed by highlighting the relationship between land use practices
and river health.

The research can add to the scientific knowledge based on the interactions
between LULC changes and river discharge, particularly in tropical watershed systems,
and help to improve the accuracy and reliability of hydrologic models in predicting
these interactions. Understanding how LULC changes affect river discharge can be
important for predicting and mitigating the risks of flooding and other water-related
disasters in the Manaba River watershed.

1.4 Scope and Limitations of the Study

This study will quantify LULC changes in the Manaba Watershed and assess
their impact on the Manaba River's discharge, focusing on the area near the Cuenco
bridge. Utilizing a hydrologic model and data from sources such as the Japan
Aerospace Exploration Agency’s (JAXA's) Real Time Rainfall Watch and National
Aeronautics and Space Administration’s (NASA's) Prediction Of Worldwide Energy
Resources (POWER) Project, it will analyze the relationships between LULC changes,
river discharge, and hydrological processes [10, 11]. Additionally, it will examine the
correlation between rainfall and discharge, aiming to provide recommendations for
water resource management in response to LULC changes in the watershed.

The study's accuracy is limited by the range of river discharge data, only
available near the Cuenco bridge, and the quality of LULC change data in the Manaba
watershed. Its scope is constrained by the specific hydrologic model used, which may
produce different results compared to other models. The study might not account for
all factors affecting the Manaba River's discharge, such as climate variations and
human activities. Additionally, the findings may not be applicable to other watersheds
with different climates, geographies, and land uses. The study also may not fully
reflect the long-term impacts of LULC changes on river discharge, as these effects
could take years to manifest.

3
 CHAPTER 2

REVIEW OF RELATED LITERATURE

INTRODUCTION

2.1 LULC Change

2.1.1. Impacts of LULC Changes on Watershed Hydrology

The effects of changes in LULC on hydrological processes in watersheds have

been extensively studied globally over the past few decades [13]. Previous studies [14,
4, 15, 16, 17] have explored the impact of LULC change on hydrological processes in
various watersheds and river basins, highlighting significant alterations in water
quantity and quality, streamflow patterns, and sediment transport. These
investigations have demonstrated the significant effects of land use changes and
overall watershed dynamics and underscored the importance of evaluating such
effects to help inform local resource management.

2.2 SWAT
2.2.1. Role of SWAT in Assessing Water Dynamics
An excellent technique for estimating LULC effects is the SWAT, a
comprehensive, continuous-time model initially crafted in the early 1990s, designed
explicitly for watersheds ranging from a few hundred to several thousand square
kilometers. Operating on a semi-distributed framework, it spatially represents specific
parameters while considering others globally. This widely acknowledged and
extensively utilized model is employed to analyze the impacts of climate, soil,
vegetation, and agricultural activities on water flow dynamics [6, 18].
Building on the foundation of established models and methods, such as the
SWAT, for understanding the intricate relationships between land use changes and
watershed hydrology, it is crucial to thoroughly examine the literature in an organized
manner. Key subtopics to cover in the literature review include previous studies on

4
LULC change and hydrology, methods and models for assessing LULC impacts on
hydrological processes, watershed management strategies, and correlation analysis of
river discharge and rainfall.

BODY

2.3 LULC
2.3.1 LULC Change and Hydrology
Land use refers to how people use the land for various socio-economic
activities such as entertainment, animal habitat, or agriculture. Land cover refers to
the material that is present on the Earth's surface, such as vegetation, urban
infrastructure, water, and bare soil. LULC is the classification of human activities and
natural components on the landscape within a certain time period based on scientific
and statistical examination of source data [4].
LULC change is the cause of ecosystem change, which alters the hydrological
process. These changes in land cover have affected the hydrological cycle and flood
vulnerability. Therefore, gaining insight into the effects of alterations in land use on
hydrology at the watershed scale can aid in the creation of sustainable water resource
strategies [16].
As remarked by Grey et al., “the primary focal point of watersheds is the river
systems” [19]. Rainfall travels through watersheds to creeks, streams, rivers,
reservoirs, bays, and eventually to the ocean. These river systems are critical because
they transport water across the landscape. Hence, rivers play a vital role in managing
watersheds, particularly in tropical island nations and other equatorial regions [19].

2.4 Methods and Models

2.4.1. Methods and Models for Assessing LULC Impacts on Hydrological Processes

The SWAT model is a hydrologic tool used to evaluate water resources and
anticipate the effects of alterations in land use, land cover, land management

5
techniques, soil erosion, sedimentation, and nonpoint source pollution in watersheds
or large river basins [6, 7, 20]. Physical algorithms are computational procedures
based on fundamental principles of physics and natural processes. They simulate the
behavior of physical systems, such as those involved in hydrology, by representing the
interactions between various environmental factors. SWAT employs physical
algorithms for runoff estimation, utilizing principles of precipitation, soil
characteristics, topography, land cover, and management practices to predict water,
sediment yield, and nonpoint source pollution in large complex watersheds [16].
The availability of land use, flow, and climate data enhances the SWAT model's
capability to understand and model changing hydrologic responses over time [14]. For
that reason, the SWAT model is a valuable tool for simulating the hydrology of tropical
island watersheds [19]. Initial findings indicate that the SWAT model is highly effective
at modeling water dynamics and sediment transport mechanisms [18]. However, it is
important to note that the simulated results, while realistic, should be considered
experimental and ongoing rather than conclusive or final [19].
Return periods, which denote the average time interval between extreme
hydrological events, play a critical role in understanding watershed behavior. By
leveraging the SWAT model, several researchers found that shorter return periods
resulted in greater alteration in peak discharge and runoff volume when compared to
longer return periods, in response to changes in land use conditions [4]. Furthermore,
Ghaffari, G., et al. (2010), investigating the effects of land-use and land-cover change
(LULC) on hydrological cycles, found that SWAT-based simulations reveal a nonlinear
and threshold-driven response to such changes [16]. This highlights the complex
relationship between land use and hydrological processes, emphasizing the need for
nuanced analysis.

Despite its strengths, the SWAT model has limitations, notably in its
representation of hydrological response units within sub-basins [3]. Unfortunately,
one such limitation is its inability to precisely represent the spatial organization of
hydrological response units within sub-basins. This means that it struggles to
accurately depict the layout of rivers, streams, and other water features within these
areas. Consequently, this deficiency could lead to inaccuracies in predicting

6
phenomena such as flooding, water quality distribution, and ecological impacts within
those regions.
Kibii, J. K., et al. (2021), utilizing the SWAT model, investigated the impact of
land use change and climate variability on catchment yield — which refers to the total
quantity of surface water expected from a stream's catchment area, found that the
alteration in hydrologic behavior is linked to the variation in forest cover. The
reduction in forest cover was considered to be likely caused by charcoal burning and
illegal logging, probably resulting in decreased water on the ground surface entering
the soil and water moving downward in the soil, through the soil and layers beneath
the surface, towards the groundwater due to reduced surface roughness. The
reduction in water transferred from the land to the atmosphere by evaporation from
the soil and other surfaces and by plant transpiration likely impacts the catchment's
hydrological cycle. Thus, these changes likely resulted in decreased precipitation, a
higher percentage of rainfall turning into runoff, and reduced groundwater recharge.
Additionally, Xu, H., et al. (2010), explored the sensitivity of hydrological
models to climate datasets, highlighting the importance of parameterization in
accurately representing relationships between groundwater and streamflow. This
sensitivity aligns with the observations made in previous studies, further emphasizing
the interconnectedness of various factors in hydrological modeling [14].

2.5 Watershed
2.5.1. Watershed Management Strategies
The findings from SWAT model applications and other studies provide critical
insights into the complex interplay between land use changes and hydrological
processes. These insights are pivotal for developing effective watershed management
strategies that aim to mitigate the adverse effects of LULC changes and promote
sustainable water resource management. A crucial element of LULC change detection,
a process of identifying differences in the state of an object or phenomenon by
observing it at different times, is identifying the specific transformation occurring, i.e.,
the conversion from one land use category to another [2]. This information will
disclose both favorable and unfavorable alterations and categories that are relatively

7
consistent over time [2]. Understanding the correlation between hydrologic response
and LULC is crucial for managing water resources and planning land use.
A study by Briones, R. U., et al. (2016), which examined the impact of LULC on
the hydrologic characteristics of a significant watershed in Batangas, Philippines,
utilizing the SWAT model, concluded that model simulations indicate that annual
surface runoff and water output increase, whereas baseflow and groundwater
recharge decrease with loss of forest cover, reduction in rangeland, and expansion of
cultivated land and urban areas. Balancing the reduction in rangeland and the
expansion of cultivated land is crucial to ensuring water security and sustainability in
the watershed and surrounding regions amidst these contrasting land use and cover
change scenarios. [15]. The outcomes obtained from the simulations can assist in
formulating and identifying effective policies to mitigate the impacts of LULC changes
on agriculture and water resources [17]. Some examples of these policies are:
programs that help sustain the watershed, ensuring that the watershed and the
surrounding regions are not full of trash, and strategies that assist in aiding the health
of the watershed.

2.5.2. Correlation Analysis Between River Discharge and Rainfall

Rainfall and river discharge are two concepts in hydrology that can influence
watershed activities. A study conducted by Hartini, S., et al. (2015), which examined
the relationship between rainfall and river discharge using the Vector Auto Regression
(VAR) model. The VAR model is a statistical model that utilizes time series data to
forecast the variable of a system. VAR has the capability to examine the dynamic
effects of disturbance factors on the system variable. It can be utilized to comprehend
the mutual connection between variables, as well as to determine the relationship
between causality and impulsive response of correlated variables, and to understand
the inter-relationship between variables [21]. Using the VAR model, the study found
that both rainfall and the initial condition of the river discharge strongly influence the
quantity of river discharge. As a result, there are auto-correlation relationships within
the river discharge [21]. However, the relationship between rainfall and streamflow,
as river discharge indicates, is not always linear [21]. Additionally, the relationships

8
between river discharge and rainfall variables indicate that rainfall in a studied region
follows a downstream pattern from its upstream origin [21]. However, this research is
limited by the use of daily data, a factor that might affect the accuracy of the forecast,
especially about the peak of the river flow due to the time lag. By utilizing this type of
data, it is presumed that the precipitation is widely spread throughout the day,
although in fact, the rain may have only fallen at specified hours each day. Similarly,
the river's discharge figures may fluctuate during the day. Enhancing prediction
accuracy may be achieved by using data gathered at shorter time intervals, such as
hourly data, instead of data collected over longer periods [21].
According to a study conducted by Tuladhar, D., et al. (2019), which examined
the long-term trends and fluctuations in the headwater, upper, and middle
catchments of the Bagmati River, there is no clear spatial pattern in the relationship
between rainfall and river discharge, which varies based on basin characteristics,
including LULC [20]. Analysis of correlation in conjunction with LULC changes also
suggested that river discharge is highly dependent on rainfall in urban areas, and the
correlation between them is weaker when the extent of human activities, especially
urbanization, is limited [20].

CONCLUSION

2.6 SWAT Applications and LULC Change Analysis

The importance of understanding LULC in ecosystem changes and their
influence on hydrological processes, particularly at the watershed scale, cannot be
overstated. Although SWAT has played a crucial role in modeling water dynamics with
varying outcomes, it's crucial to acknowledge that its results should complement the
simulation of LULC changes rather than serve as definitive evidence. Nonetheless,
several studies have demonstrated that SWAT simulations can contribute to
watershed management efforts and aid in mitigating the impacts of LULC changes. For
example, a study by Briones, R. U. et al. (2016) utilized SWAT simulations to assess the
impact of LULC changes on hydrologic response of Palico Watershed using the SWAT
model [15]. Similarly, Chaemiso, S. E. et al. (2021) applied SWAT models to evaluate
9
the effect of LULC changes on surface water availability in the Omo-Gibe Basin,
Ethiopia. These examples illustrate the practical utility of SWAT in managing
watersheds and addressing the challenges posed by LULC changes [17].

The SWAT model developed by Srinivasan, R., et al. (2010) is widely regarded
as a dependable baseline model for various analysis scenarios, offering unbiased
results. This review emphasizes the effectiveness of spatially distributed models in
producing satisfactory outcomes using readily available input parameters, applicable
to watersheds of varying sizes. Moreover, a study by Ghaffari, G. et al. (2010) analyzed
that there is recognition of the potential for further comprehensive evaluation and
enhancement by incorporating land use, flow, and climate data, making it essential
for assessing environmental factors in watersheds and river basins. Additionally, the
integration of SWAT is beneficial as it represents the interactions between various
environmental factors [16].
Ensuring the management of watersheds is paramount for maintaining
ecosystem health and safety. LULC plays a pivotal role in shaping water sustainability
within the watershed. Studies underscore the significance of understanding the
correlation between rainfall and river discharge for effective water resource
management. However, limitations in data accuracy due to time delays in collection
highlight the need for caution in interpreting research findings. Moreover, the
relationship between rainfall and river discharge varies depending on basin
characteristics. Notably, river discharge exhibits a strong dependence on rainfall, with
a weaker correlation observed in regions with fewer anthropogenic activities.
The literature review has a constraint, that being data accuracy, which could
affect the precision of the hydrological model, by inputting inaccurate data onto the
model. There is a lack of historical land-use maps for the Manaba Watershed to track
changes over an extended period and compare with simulated model changes. Using
SWAT, we aim to simulate the hydrological model of the Manaba Watershed in Garcia-
Hernandez, Bohol. The study's scope and data collection will be limited to the vicinity
of the Cuenco Bridge. This study aims to provide insights into the effects of human-
induced changes and will guide water resource management strategies for the benefit
of the river's ecosystems and surrounding communities.

10
 CHAPTER 3
MATERIALS AND METHODS

3.1 Study Design

3.1.1. Objectives of the Study

The study aims to calibrate and validate the SWAT model to ensure its accuracy
in simulating watershed properties. The objectives include a detailed simulation of
hydrological and environmental properties of a watershed from 2017 – 2022 to
analyze the effects of LULC changes. Additionally, the study seeks to understand the
correlation between rainfall events and river discharge rates, thus assessing the
watershed's response to precipitation. These steps are crucial for predicting future
watershed behaviors under varying environmental conditions.

3.1.2. Calibration and Validation

Initially, the SWAT model will be calibrated and validated to ensure its
reliability for simulating the watershed properties accurately. This step involves fine-
tuning the model parameters to align with historical data, thus ensuring that the
model can faithfully reproduce observed watershed behavior.

3.1.3. Simulation and LULC Change Analysis

After utilizing the calibrated SWAT model, simulation can now begin on the
watershed's hydrological and environmental properties throughout 2017 – 2022.
Following the simulation, an in-depth analysis will be conducted to ascertain the
effects of LULC changes throughout 2017 to 2022 on the watershed's properties. This
step involves comparing sediment yield, river discharge, suspended load, suspended
sediment concentration, surface runoff, nutrient loading, and streamflow for the

11
years, and highlighting the variations in watershed characteristics attributable to LULC
alterations.

3.1.4. Correlation Analysis

An essential component of the study involves examining the correlation

between rainfall events and river discharge within the watershed. By analyzing the
simulation results, the study aims to understand how rainfall patterns influence river
discharge rates, thereby assessing the watershed's hydrological response to
precipitation.

While the study does not explicitly state an expectation for the correlation to
change or remain the same during these years, the aim to understand how rainfall
patterns influence river discharge suggests that the study is open to discovering
variations in this relationship. The inclusion of multiple years in the analysis inherently
acknowledges the possibility of differing correlations due to year-to-year climatic
variability, changes in land use, and other environmental factors. Therefore, while
there is no stated expectation of difference, the study design implies that detecting
any variations in correlation over time is within the scope of the research objectives.

3.2 Data

3.2.1. Study Area

The Manaba Watershed, located at the southeastern part of Bohol,

Philippines, encompasses four municipalities, namely Garcia Hernandez, Jagna, Sierra
Bullones, and Valencia (Fig. 3.1). The major part of the Manaba Watershed can be
found in Garcia Hernandez. The total area of the watershed is about 5,823.77 x 10⁵
m. Located within the watershed is the Manaba River and its tributaries. The major
river system spans from the river mouth located at 1062857°N latitude – 642273°E
longitude to its headwaters at 1073982°N latitude – 640588°E longitude (Fig. 3.2). The
main river segment spans a distance of 18.52 km, with outlet widths of up to 52 m.
12
The midstream part has a width of 53 m, while the upstream section has a width of 28
m. The average width of the river is 44.33 m (Fig. 3.3). Our study area for the Manaba
Watershed is limited at the Manaba River in its downstream area, specifically at
9.6158°N latitude and 124.3005°E longitude. We mainly chose this specific location
because the Manaba Watershed has multiple sub watersheds, streams, and rivers and
we want to know if LULC changes could affect the water quantity and quality of the
Manaba River.

Figure 3.1. Watershed map of Garcia Hernandez. (Source: Municipal Planning and
Development Coordinator (MPDC) of Garcia Hernandez)

13
Figure 3.2. Coordinated map of Manaba River. (Source: R.I.V.E.Rs for Life Program of
CENTRO Tagbilaran City, 2019 [23])

14
Figure 3.3. Length map and width map of Manaba River. (Source: R.I.V.E.Rs for Life
Program of CENTRO Tagbilaran City, 2019 [23])

3.2.2. Data Collection

The SWAT Model requires both geospatial and climatological input data in
order to begin simulating the watershed. The geospatial inputs that are needed are
topographic map, LULC map, and soil map. Meanwhile, the climatological input data
that are needed are temperature, rainfall, wind, relative humidity, and solar radiation
data. The topographic map can be obtained from Advanced Land Observing Satellite
(ALOS) Digital Surface Model (DEM), the LULC map from the European Space Agency
(ESA) Sentinel-2, and the soil map from the Food and Agriculture Organization
(FAO) Digital Soil Map of the World. On the other hand, the rainfall data can be
obtained from JAXA Realtime Rainfall Watch, while the and temperature, wind,
relative humidity, and solar radiation data from the NASA Langley Research Center
POWER Project. Additionally, river discharge data can be obtained from the MPDC of
Garcia Hernandez and Department of Public Works and Highways (DPWH), specifically
from staff gages located under the Cuenco Bridge. All input data for the SWAT

15
simulations mentioned above were gathered from available global and local datasets.
Table 3.1 lists the types of data and their corresponding sources.

Table 3.1. Model inputs and data sources for the SWAT simulations.
Data Source Resolution Period
Geospatial
2006 –
DEM ALOS 30 m
2011
Soil map FAO - 2007
2007 and
LULC maps ESA Sentinel-2 L2A (2017 and 2022) 10 m
2023
Climatology
Rainfall
Temperature
2012 -
Wind speed NASA POWER Project -
2023
Solar radiation
Relative
humidity

The LULC data will be in a bit-map graphics format, representing different land
cover types across the watershed. Rainfall data will be in time-series format, providing
precipitation measurements over the study period. Furthermore, the soil and
topographic data will also be similar to the format of the LULC data, which helps in
detailing the physical characteristics of the watershed.

16
3.3 Model

3.3.1. Model Flowchart

Figure 3.4 shows the flowchart of the SWAT model to be used in this study.
Required data are geospatial and climatological. In this scenario modeling study,
different LULC maps will be used in separate simulations to investigate the impacts of
LULC change.

Figure 3.4. SWAT model flowchart.

3.3.2. Governing Equation

The governing equation used by SWAT for sedimentation is the Modified

Universal Soil Loss Equation (MUSLE) developed by Williams (1975) [8] (equation 1).
17
 Sed = 11.8(QsurfqpeakAhru)EIUSLEKUSLECUSLEPUSLELSUSLECFRG (1)
where
Sed = sediment yield on a given temporal scale (t ha−1 )
Qsurf = surface runoff volume (mm h−1 )
qpeak = peak runoff rate (m3 s−1)
Ahru = area of HRU (ha)
KUSLE = soil erodibility factor of USLE
CUSLE = cover and management factor of USLE
PUSLE = support practice factor of USLE
LSUSLE = topographic factor of USLE
CFRG = coarse fragment factor.

The MUSLE is a modified version of the Universal Soil Loss Equation (USLE)
which was developed by Wischmeier and Smith (1965). While the USLE uses rainfall
as an indicator of erosive energy, MUSLE uses the amount of runoff to simulate
erosion and sediment yield. For a detailed rundown of the sediment modeling
component of SWAT and its validity, refer to the work of Neitsch et al. (2001) [25].

3.3.3. Model Input

DEM

A DEM with a 30m resolution will be used, as opposed to counterparts with

90m resolution. It was clipped from a global dataset (AW3D30) generated from images
collected using the Panchromatic Remote-sensing Instrument for Stereo Mapping
(PRISM) aboard ALOS from 2006 to 2011 [10] (Fig. 3.5).

18
Figure 3.5. DEM of Bohol region around Manaba watershed. (Source: Esri, Maxar,
Earthstar Geographics, and the GIS User Community)

The DEM will be used by SWAT to generate the stream network, delineate the
watershed, and create a slope map. The latter will be overlaid with the soil and LULC
maps to define Hydrologic Response Units (HRUs) in the model domain. A maximum
number of 5 user-defined slope classes can be configured in SWAT. Figure 3.6 shows
the map with these slope classes in Manaba watershed. Slope classes defined were
followed (Table 3.2) [13].

19
Figure 3.6. Slope map of Manaba watershed generated from the DEM. (Source: Esri,
Maxar, Earthstar Geographics, and the GIS User Community)

Table 3.2. User-defined slope classes configured in the SWAT set-up. (Source: 13)

Slope range (%) Description

0–3 Flat
3–6 Rolling
6 – 15 Hilly
15 – 30 Steep
>30 Very steep

Upon communication with the Bureau of Soils and Water Management

(BSWM) requesting for a high-resolution local soil map, the Soil Survey Division is still

20
in the process of converting the analog soil datasets and maps into digital compilation
that would be readily useful for integration into any spatial database system. Hence,
the coarser Digital Soil Map of the World (DSMW) will be used (Fig. 3.7) [26]. The
downloaded shapefile will be rasterized on ArcGIS before introducing it to the SWAT
model.

Figure 3.7. Digital soil map of the world (Source: 26)

21
LULC Maps

LULC maps with 10m resolution from ESA Sentinel-2 L2A imagery are provided
by Impact Observatory, Microsoft, and Esri [28]. This resolution is remarkably better
compared to other considered sources such as the Dynamic Land Cover map (CGLS-
LC100) of the Copernicus Global Land Service which has a 100 m resolution [24].
Another considered data source was from the GlobeLand30, a 30m resolution global
land cover data product for 2000, 2010, and 2020 developed by China [29]. This would
have been a fitting source to investigate the decadal changes in LULC, but upon closer
examination, inaccuracies that oppose expected trends were observed. For one,
decreased urban area coverage was mapped in their 2020 product compared to 2010
and 2000 when historical satellite images clearly indicate urban expansion as the years
progressed.

The earliest LULC map version from the most suitable source which is that from
ESA Sentinel-2 is 2017. The product is generated with Impact Observatory’s deep
learning AI land classification model, trained using billions of human-labeled image
pixels from the National Geographic Society. They provide annual versions of this
product and the latest version as of date is 2023. The earliest and latest versions were
chosen; hence, this study investigates the effect of a 6-yr change in LULC in Manaba
watershed. Values were reclassified into land use classes that can be read by SWAT
(Table 3.3).

Table 3.3. Look-up table used in SWAT on the ESA Sentinel-2 LULC maps.

Value Land Use Code Description

1 WATR Water
2 FRST Forest

22
 3 WETL Wetland
4 AGRL Agriculture
5 URHD Urban Area
6 BARR Barren land
7 RNGE Grassland
8 RNGB Shrubland

Climatology

Daily weather data for rainfall, temperature (maximum and minimum), wind
speed, solar radiation, and relative humidity will be obtained from the
agroclimatology product of the POWER Project by NASA Earth Science’s Applied
Sciences Program. Their meteorological data/parameters were based upon the
Goddard’s Global Modeling and Assimilation Office (GMAO) Modern Era
Retrospective-Analysis for Research and Applications (MERRA-2) assimilation model
products and GMAO Forward Processing – Instrument Teams (FP-IT) GEOS 5.12.4
near-real time products. Their solar based data/parameters were based upon satellite
observations with subsequent inversion to surface solar insolation by NASA’s Global
Energy and Water Exchange Project (GEWEX) / Surface Radiation Budget (SRB) Release
3 and NASA’s CERES Fast Longwave and Shortwave Radiative project (FLASHFlux) [11].
Figure 3.8 shows the user-defined weather station location in Manaba watershed.

23
Figure 3.8. Weather station location (Latitude: 9°41'4.97"N; Longitude:
124°17'49.03"E; elevation: 335 m) in the study domain. (Source: Esri, Maxar, Earthstar
Geographics, and the GIS User Community)

3.4 Model Output

3.4.1. Observed Data for Model Calibration and Validation

Calibration and validation in hydrologic modeling studies are imperative

whenever feasible. Historical streamflow from the Streamflow Management System
database of the Department of Public Works and Highways - Bureau of Design - Water
Projects Division [30] for a river gauge located in Purok 1, Brgy. Calma, Garcia-
Hernandez, Bohol (Latitude: 9°37'50"N; Longitude: 124°18'16"E, Figure 3.9) will be
collected for use in the calibration and validation of our SWAT model.

24
Figure 3.9. River gauge location to be used for model calibration and validation.
(Source: Esri, HERE, Garmin, Intermap, increment P Corp., GEBCO, USGS, FAO, NPS,
NRCAN, GeoBase, IGN, Kadaster NL, Ordinance Survey, Esri Japan, METI, Esri China
(Hong Kong), (c) OpenStreetMap contributors, and the GIS User Community)

3.4.2. Model calibration and validation

Calibration will be done using SWAT-CUP2019. This open-source software links

SUFI2, GLUE, ParaSol, MCMC, and PSO algorithms to SWAT. The most widely used and
accepted SUFI2 algorithm will be used. The SUFI2 algorithm maps all uncertainties on
the parameters and tries to capture the measured data within the 95% prediction
uncertainty (95PPU) of the model in an iterative process [31]. Fitted parameters in the
model are determined by the calibration tool in order to generate values that are close
to actual values. Time scale in determining fitted parameters is monthly as
recommended [7] and adapted by most SWAT studies.

Five hundred (500) iterations will be done in SWAT-CUP and the iteration with
the best statistical results will be chosen. Statistical tools used to evaluate model
performance are the coefficient of determination (𝑅 2 ) [32], the Nash-Sutcliffe model

25
efficiency (𝑁𝑆𝐸) [27, 28], and the percentage deviation (𝑃𝐵𝐼𝐴𝑆) [31]. The calculation
formulas are as follows:

(2)

(3)

(4)

where 𝑄𝑚 is the measured data, 𝑄̅𝑚 is the mean of the measured data, 𝑄𝑠 is the model
simulation data, and 𝑄̅𝑠 is the mean of the model simulation data.
The 𝑅 2 indicates the degree of linear relationship between simulated and
observed data. An 𝑅 2 value close to one ideally indicates a better performance.
However, it is very sensitive to extremely high values and not indicative enough in
determining the adequacy of a model. One can have a low 𝑅 2 value for a good model,
or a high 𝑅 2 value for a model that does not fit the data. The 𝑁𝑆𝐸 is one of the most
commonly used criteria and is widely applied in hydrologic models. This is a
normalized statistic which can be used to determine the goodness of fit. It ranges from
negative infinity to 1, with 1 being the ideal value [33]. The 𝑃𝐵𝐼𝐴𝑆 assesses the
average deviation of the simulated values from the observed values, with 0 as the
ideal value, and a positive (negative) 𝑃𝐵𝐼𝐴𝑆 value representing an underestimation
(overestimation) bias of the simulated variable compared to the measured variable
[35]. A model is considered satisfactory if it satisfies the following criteria: 𝑅 2 > 0.6;
𝑁𝑆𝐸 > 0.5; 𝑃𝐵𝐼𝐴𝑆 < ± 25 [33, 36].

3.4.3. Correlation Analysis Equation

26
 In determining the relationship between river discharge and rainfall, we will
employ the Pearson Product Moment Coefficient of Correlation, r. To compute r, we
can use the formula (other forms can be derived from this) [37] (equation 5).

n ΣXY − ΣX ∙ ΣY
r=
√[n ΣX 2 − (ΣX)2 ][n ΣY 2 − (ΣY)2 ]
(5)

where

n = sample size

X = rainfall

Y = river discharge

The direction of the correlation is determined by the sign of the coefficient r,

which can be either positive or negative. The magnitude of correlation is determined
by taking the absolute values of r. Table 3.4 interprets the values of r [37].

Table 3.4. Pearson correlation table

Pearson 𝐫 Qualitative description

1.00 Perfect
0.75 ≤ r < 1.00 Very high
0.5 ≤ r < 0.75 Moderately high
0.25 ≤ r < 0.5 Moderately low
0.0 < r < 0.25 Very low
0.00 No correlation
Source: [37]

27
3.4.5. Model Runs
There will be two separate SWAT simulations - one using LULC 2017 (SIM2017)
and another using LULC 2023 (SIM2023). The model will be made to run for 12 years
(2012 – 2023), but the first 3 years will be set as the warmup period so the results to
be discussed will only be from 2015 to 2023. Discussions will focus on the parameters
listed in Table 3.5.

Table 3.5. Output parameters.

Parameter SWAT Code Unit Description
Basin Values
Amount of sediment
Sediment yield SED YIELD 𝑡⁄ leaving the watershed in a
ℎ𝑎
given period of time.
Unconfined water over the
Surface runoff SURF_Q 𝑚𝑚
ground surface.
Amount of groundwater
Baseflow GW_Q 𝑚𝑚
flowing into the channel.
Water transferred from the
land to the atmosphere by
Evapotranspiration ET 𝑚𝑚 evaporation from surface
and transpiration from
plants.
River Mouth Values
The volume of water
Streamflow FLOW_OUT 𝑚3⁄ flowing through the river
𝑠
channel per second.

28
 Sediment transported with
Suspended load SED_OUT 𝑀𝑒𝑡𝑟𝑖𝑐 𝑡𝑜𝑛𝑠
water during time step.
Suspended sediment 𝑚𝑔 Concentration of sediment
SED_CONC ⁄𝐿
concentration during time step.

3.4.6. Evaluating the Effects of LULC Change

As an attempt to quantify the effect of the 6-year change in LULC on the

watershed and river mouth processes, the difference of values (SIM2023 minus SIM2017)
for each parameter will be calculated.

3.4.7. Test for Statistical Significance

Paired sample t-test will be employed on the model outputs. A t-test is a

statistical test that is used to compare the values of two groups and determine
whether there is significant difference in the values [38]. In this study, satisfying the
test (p-value < 0.05) means that the changes applied on the model input (hence, LULC
change) have significant effects on the hydrologic processes that are simulated.

3.4.8. Model Limitations

Since a single Digital Elevation Map and Soil Map will be used in all simulations,
changes in topography and soil compositions within the watershed are unaccounted.
In a related manner, tectonic activities cannot be considered by the model. River
mouth migration cannot also be accounted for.

LULC classes provided by the source are limited. The model cannot resolve
finer details in land use such as distinction of high-density urban areas (URHD) from
medium-density (URMD) and med/low (URML) areas. Effects of mining/quarry areas
cannot be assessed since the dataset does not distinguish such LULC class.
29
 There is a period mismatch on the geospatial data to be used in the model
runs. DEM was generated from a collection of images from 2006 to 2011. The most
recent version of the Digital Soil Map of the World was in 2007, although it is likely
that only minor changes happened in the actual topography and soil composition
through the decades. LULC is relatively fast-changing but LULC maps that can be used
are also only static (2017 and 2023 versions). SWAT simulations will be made to run
from 2012 to 2023. Hence, it is advised to view this study as merely scenario modeling.

Climatology distribution within the watershed is non-uniform in actuality (e.g.,

due to orographic effects), but since we only have one weather station, this non-
uniformity cannot be accounted for by the model.

Due to uncertainties whether the model can generate accurate values on a

daily basis, temporal resolution of the model set-up is monthly. Values declared by
the model are monthly-averaged and are not expected to match reported values in
news articles during extreme discharge events which usually last for only several
hours.

30
 LITERATURE CITED

[1] Singh, Y. (2018). Significance of Land Use / Land Cover Maps. Strategic Advanced
Technology Products and Linkages Development Agency (SATPALDA).
Retrieved from https://www.satpalda.com/blogs/significance-of-land-use-
land-cover-lulcmaps.
[2] Zubair, A. O. (2006). Change detection in land use and land cover using Remote
Sensing Data and GIS (A case study of Ilorin and its environs in Kwara State
Nigeria). MSc Thesis. University of Ibadan. Retrieved from
https://www.geospatialworld.net/wpcontent/uploads/2016/04/OpeyemiZub
air_ThesisPDF.pdf.
[3] Aich, V., Liersch, S., Vetter, T., Fournet, S., Andersson, J. C. M., Calmanti, S., van
Weert, F. H. A., Hattermann, F. F., Paton, E. N. (2016). Flood projections within
the Niger River Basin under future land use and climate change. Science of The
Total Environment, 562, 666–677. DOI:
https://doi.org/10.1016/j.scitotenv.2016.04.021.
[4] Zope, P. E., Eldho, T. I., & Jothiprakash, V. (2016). Impacts of land use–land cover
change and urbanization on flooding: A case study of Oshiwara River Basin in
Mumbai, India. CATENA, 145, 142–154. DOI:
https://doi.org/10.1016/j.catena.2016.06.009.
[5] Stewart, R. H. (2000). Introduction to Physical Oceanography. Department of
Oceanography, Texas A & M University.
https://www.uv.es/hegigui/Kasper/por%20Robert%20H%20Stewart.pdf?she
msssxt.
[6] Arnold, J. G., & Fohrer, N. (2005). SWAT2000: Current capabilities and research
opportunities in applied watershed modelling. Hydrological Processes, 19(3),
563–572. https://doi.org/10.1002/hyp.5611.
[7] Srinivasan, R., Zhang, X., & Arnold, J.G. (2010). SWAT Ungauged: Hydrological
Budget and Crop Yield Predictions in the Upper Mississippi River Basin.

31
 Transactions of the ASABE, 53, 1533-1546.
https://doi.org/10.13031/2013.34903.5.
[8] Williams, J. R. (1975). Sediment routing for agricultural watersheds. Water Resour.
Bull. Am. Water Resour. Assoc., 11 (5) (1975), pp. 965-974.
https://doi.org/10.1111/j.1752-1688.1975.tb01817.x.
[9] DENR (2019). Bohol’s Manaba River wins as country’s 3rd cleanest river
BOHOLCHRONICLE.COM.PH. Retrieved from
https://www.boholchronicle.com.ph/2019/06/19/bohols-manaba-river-
winsas-countrys-3rd-cleanest-river-denr.
[10] JAXA Realtime Rainfall Watch. (2021). Sharaku.eorc.jaxa.jp.
https://sharaku.eorc.jaxa.jp/GSMaP_NOW/. (Accessed 2023)
[11] NASA Langley Research Center (LaRC). Prediction of Worldwide Energy Resources
(POWER) Project. NASA Earth Science/Applied Science Program. Retrieved
from https://power.larc.nasa.gov/. (Accessed 2023)
[12] Provincial Planning and Development Office of Bohol. (2023). Watershed map.
Retrieved from https://ppdo.bohol.gov.ph/maps/thematic-
maps/watershedmap/.
[13] Boongaling, C. G. K., Faustino-Eslava, D. V., & Lansigan, F. P. (2018). Modeling land
use change impacts on hydrology and the use of landscape metrics as tools for
watershed management: The case of an ungauged catchment in the
Philippines. Land Use Policy, 72, 116–128.
https://doi.org/10.1016/j.landusepol.2017.12.042.
[14] Kibii, J. K., Kipkorir, E. C., & Kosgei, J. R. (2021). Application of Soil and Water
Assessment Tool (SWAT) to Evaluate the Impact of Land Use and Climate
Variability on the Kaptagat Catchment River Discharge. Sustainability, 13(4),
1802. https://doi.org/10.3390/su13041802.
[15] Briones, R. U., Ella, V. B., & Bantayan, N. C. (2016). Hydrologic Impact Evaluation
of Land Use and Land Cover Change in Palico Watershed, Batangas, Philippines
Using the SWAT Model. Journal of Environmental Science and Management,

32
 19(1), 96–107. https://ovcre.uplb.edu.ph/journals-
uplb/index.php/JESAM/article/download/ 159/141.
[16] Ghaffari, G., Keesstra, S. D., Ghoddousi, J., & Ahmadi, H. (2010). SWAT-simulated
hydrological impact of land-use change in the Zanjanrood basin, Northwest
Iran. Hydrological Processes, 24(7), 892–903.
https://doi.org/10.1002/hyp.7530.
[17] Chaemiso, S. E., Kartha, S. A., & Pingale, S. M. (2021). Effect of land use/land cover
changes on surface water availability in the Omo-Gibe Basin, Ethiopia.
Hydrological Sciences Journal, 66(13). 1936-1962.
https://doi.org/10.1080/02626667.2021.1963442.
[18] Hallouz, F., Meddi, M., Mahé, G., Alirahmani, S., & Keddar, A. (2018). Modeling of
discharge and sediment transport through the SWAT model in the basin of
Harraza (Northwest of Algeria). Water Science, 32(1), 79–88.
https://doi.org/10.1016/j.wsj.2017.12.004 9.

[19] Grey, O. P., Webber, D. F. S. G., Setegn, S. G., & Melesse, A. M. (2014). Application
of the Soil and Water Assessment Tool (SWAT Model) on a small tropical island
(Great River Watershed, Jamaica) as a tool in Integrated Watershed and
Coastal Zone Management. Revista de Biología Tropical, 62(Suppl. 3), 293–305.
https://www.scielo.sa.cr/scielo.php?script=sci_arttext&pid=S0034-77442014
000700004.
[20] Tuladhar, D., Dewan, A., Kuhn, M., & Corner, R. J. (2019). The Influence of Rainfall
and Land Use/Land Cover Changes on River Discharge Variability in the
Mountainous Catchment of the Bagmati River. Water, 11(12), 2444.
https://doi.org/10.3390/w11122444.
[21] Hartini, S., Hadi, M. P., Sudibyakto, S., & Poniman, A. (2015). Application of Vector
Auto Regression Model for Rainfall-River Discharge Analysis. Forum Geografi,
29(1), 1–10. https://journals.ums.ac.id/index.php/fg/article/view/786/515.
[22] Douglas-Mankin, K. R., Srinivasan, R., & Arnold, J. G. (2010). Soil and Water
Assessment Tool (SWAT) Model: Current Developments and Applications.

33
 Transactions of the ASABE, 53(5), 1423–1431.
https://doi.org/10.13031/2013.34915.
[23] R.I.V.E.Rs for Life Program of CENRO Tagbilaran City (pp. 1–14). (2019).
DEPARTMENT OF ENVIRONMENT AND NATURAL RESOURCES COMMUNITY
ENVIRONMENT AND NATURAL RESOURCES OFFICE.

[24] Wischmeier, W.H. and Smith, D.D. (1965). Predicting rainfall-erosion losses from
cropland east of the Rocky Mountains. Agriculture Handbook 282. USDA-ARS.
19.

[25] Neitsch, S.L, Arnold, J. G., Kiniry, J. R., & Williams, J. R. (2001). Soil and water
assessment tool user’s manual. Version 2000. October 1999. Grassland, Soil
and water research laboratory.

[26] Food and Agriculture Organization of the United Nations (FAO-UNESCO) (2007).
Digital Soil Map of the World. Retrieved from
https://data.apps.fao.org/map/catalog/static/api/records/446ed430-8383-11
db-b9b2-000d939bc5d8.

[27] Buchhorn, M., Lesiv, M., Tsendbazar, N., Herold, M., Bertels, L., Smets, B. (2020).
Copernicus Global Land Cover Layers-Collection 2. Remote Sensing, 12 Volume
108, 1044.

[28] Karra, K., Kontgis, C., Statman-Weil, Z, Mazzariello, J. C., Mathis, M., & Brumby, S.
(2021) Global land use / land cover with Sentinel 2 and deep learning. IEEE
International Geoscience and Remote Sensing Symposium IGARSS, Brussels,
Belgium, 2021, pp. 4704-4707.

[29] Jun, C., Ban, Y. & Li, S. Open access to Earth land-cover map. Nature 514, 434
(2014).

[30] NHDCP (2016). National Hydrologic Data Collection Program. Streamflow

Management System, Department of Public Works and Highways – Bureau of

34
 Design – Water Projects Division. Retrieved from
https://apps.dpwh.gov.ph/streams_public/station_public.aspx.

[31] Abbaspour, K.C., Rouholahnejad, E., Vaghefi, S., Srinivasan, R., Yang, H., and Klove,
B. (2015). A continental-scale hydrology and water quality model for Europe:
calibration and uncertainty of a high resolution large-scale SWAT model.
Journal of Hydrology, 524 (2015), 733–752.

[32] Krause, P., Boyle, D. P., & Bäse, F. (2005). Comparison of different efficiency
criteria for hydrological model assessment. Advances in Geosciences, 5, 89–
97.

[33] Moriasi, D. N., Arnold, J. G., Liew, M. W. Van, Bingner, R. L., Harmel, R. D., & Veith,
T. L. (2007). Model Evaluation guidelines for systematic quantification of
accuracy in watershed simulations, 50(3), 885–900.

[34] Nash, J. E., & Sutcliffe, J. V. (1970). River flow forecasting through conceptual
models part I — A discussion of principles. Journal of Hydrology, 10(3), 282–
290. 20.

[35] Gupta, H. V., Sorooshian, S., & Yapo, P. O. (1999). Status of automatic calibration
for hydrologic models: comparison with multilevel expert calibration. Journal
of Hydrologic Engineering, 25(11), 1093–1120.

[36] Neupane, R.P., Kumar, S., (2015). Estimating the effects of potential climate and
land use changes on hydrologic processes of a large agriculture dominated
watershed. J. Hydrol. 529, 418– 429.

[37] Belecina, R. R., Baccay, E. S., & Mateo, E. B. (2016). Statistics and Probability (First
Edition, pp. 282–303). REX Book Store.

[38] Bevans, R. (2022). An Introduction to t Tests. Definitions, Formula and Examples.

Scribbr. Retrieved May 21, 2023, from https://www.scribbr.com/statistics/t-
test/

35
36