INTERNATIONAL JOURNAL OF MULTIDISCIPLINARY RESEARCH AND ANALYSIS

ISSN(print): 2643-9840, ISSN(online): 2643-9875

Volume 07 Issue 00 Month 2024
DOI: 10.47191/ijmra/v7-i0-00, Impact Factor: 8.22
Page No. 00

Cauayan City Flood Susceptibility

Mapping using GIS and Remote Sensing

Author: Ador Y. Franco MIT Co-Author: Dr. Arnel C. Fajardo

Isabela State University Cauayan Campus Isabela State University Cauayan Campus
E-mail: [email protected] E-mail: [email protected]

ABSTRACT: Floods are among the most destructive natural disasters, causing extensive damage to life, property, and the
environment. Flood susceptibility mapping (FSM) plays a critical role in mitigating flood risks and managing disaster response.
This study explores the use of Geographic Information Systems (GIS) and Remote Sensing (RS) technologies to create accurate
and reliable flood susceptibility maps. By integrating spatial data and hydrological models, this paper demonstrates how GIS and
RS can identify areas vulnerable to flooding, aiding in proactive disaster planning and resource allocation.

KEYWORDS: Flood, GIS, Machine Learning, Remote Sensing, Mapping, Flood Susceptibility Mapping

I. INTRODUCTION

Flood susceptibility mapping is a critical tool for understanding flood risks, particularly in areas prone to frequent
flooding or extreme weather events. With the increasing frequency and intensity of flooding worldwide due to climate
change, urbanization, and deforestation, effective flood risk assessment is crucial for planning and disaster management
(Williams & Shaw, 2019). Traditionally, flood susceptibility mapping relied on simple topographical maps and historical
flood data. However, advances in Geographic Information Systems (GIS) and Remote Sensing technologies have
revolutionized this field, providing high-resolution, real-time data that can more accurately predict flood-prone areas.

These technologies allow for the integration of various environmental, hydrological, and societal data to develop
comprehensive flood risk models. For instance, remote sensing tools like satellite imagery and LiDAR data enable the
monitoring of large-scale environmental changes, such as land use shifts and water body dynamics (García et al., 2022).
Coupled with GIS, these tools can model flood susceptibility by analyzing variables such as elevation, land cover, and
rainfall patterns (Zhang et al., 2021). More recently, machine learning techniques like Random Forest (RF) and Support
Vector Machines (SVM) have been integrated into these models, enhancing their predictive accuracy (García et al., 2022).

The combination of GIS, remote sensing, and advanced computational methods offers significant improvements over
traditional flood prediction techniques. By facilitating more accurate and timely flood susceptibility maps, these tools are
critical for disaster preparedness, urban planning, and climate change adaptation (Ahmad et al., 2022). However,
challenges remain in data integration, uncertainty quantification, and model validation, particularly in developing regions
with limited access to comprehensive datasets (Muthukumar et al., 2020). Despite these challenges, the growing body of
research underscores the importance of these technologies in mitigating the impacts of flooding and supporting sustainable
development in flood-prone areas.

Flooding poses significant challenges to sustainable development and disaster management worldwide. Factors such as
climate change, urbanization, and deforestation exacerbate the frequency and severity of floods. Traditional methods of
flood assessment often lack spatial precision, making GIS and RS essential tools for modern flood susceptibility mapping.

This paper discusses:

IJMRA, Volume 7 Issue (Month) 2024 www.ijmra.in Page 1

Cite as :Paper Title; Vol. …|Issue ….|Pg:102-107 [Pick the date]
1. The principles and methodologies behind FSM using GIS and RS.
2. Case studies demonstrating the application of these technologies.
3. The potential for integrating machine learning techniques.

II. LITERATURE REVIEW

Flood Susceptibility Mapping using GIS and Remote Sensing

Flood susceptibility mapping is an essential tool for identifying areas prone to flooding and for aiding in flood risk management.
The integration of Geographic Information Systems (GIS) and remote sensing has significantly advanced flood risk
assessment, providing higher accuracy and efficiency in predicting flood-prone zones. This literature review examines key studies
in the field, highlighting the methodologies, advancements, and challenges in flood susceptibility mapping using these
technologies.

GIS and remote sensing technologies allow for the collection and analysis of spatial and temporal data, providing valuable
insights for flood susceptibility mapping. Akinyemi et al. (2020) employed GIS-based Frequency Ratio (FR) and remote
sensing data to map flood-prone areas in Lagos, Nigeria. The study identified topographic, land-use, and hydrological data as
key variables influencing flood susceptibility. Similar studies by Nasir et al. (2023) applied an integrated GIS and remote sensing
approach in the Swat River Basin, Pakistan, using satellite imagery and digital elevation models (DEMs) to create flood risk
maps.

These studies demonstrate the ability of GIS and remote sensing to provide high-resolution data, which enhances the accuracy of
flood susceptibility mapping by analyzing factors such as elevation, slope, proximity to water bodies, and land cover changes
(Zhang et al., 2021). Remote sensing techniques like LiDAR and satellite imagery also enable real-time monitoring of flood
events and environmental changes, further improving flood prediction accuracy (García et al., 2022).

Incorporating machine learning techniques into flood susceptibility mapping has proven to enhance predictive accuracy. García et
al. (2022) employed Random Forest (RF) and Support Vector Machines (SVM) models in urban flood risk assessments,
combining these algorithms with remote sensing data. These models outperform traditional statistical methods in handling large
datasets and capturing complex, non-linear relationships between flood risk factors. Similarly, Ahmad et al. (2022) used Analytic
Hierarchy Process (AHP) integrated with remote sensing to assess flood vulnerability in the Kosi River Basin, utilizing multi-
criteria decision analysis (MCDA) to weigh the influence of factors such as rainfall, land use, and soil type.

Machine learning techniques like neural networks and deep learning algorithms are also gaining traction in flood susceptibility
studies due to their ability to process complex data and improve model accuracy (Cai et al., 2021). These methods allow
researchers to incorporate diverse data sources such as social, economic, and environmental factors into flood risk models,
providing more holistic flood assessments.

The accuracy of flood susceptibility maps is a critical concern, as inaccurate models can lead to poor decision-making in flood
risk management. To address this, many studies have focused on model validation and accuracy assessment using historical flood
data and ground-truthing methods. Muthukumar et al. (2020) evaluated the effectiveness of GIS-based flood models in
Chennai, India, by comparing model outputs to observed flood data. Performance metrics such as ROC curves, accuracy indices,
and confusion matrices were used to quantify model precision. These methods are essential for ensuring that flood susceptibility
models provide reliable and actionable results.

One of the growing concerns in flood susceptibility mapping is the impact of climate change on flood risk. Changes in rainfall
patterns, sea level rise, and increased frequency of extreme weather events exacerbate flooding in many regions (Williams &
Shaw, 2019). Zhang et al. (2021) explored the integration of climate change projections with flood susceptibility models in the
Yangtze River Basin, emphasizing the need for dynamic flood models that adapt to changing climatic conditions. This
adaptation involves using future climate scenarios to predict how shifting rainfall patterns and temperature changes may affect
flood-prone areas.

Despite the advancements in flood susceptibility mapping, several challenges remain. One of the primary limitations is the
availability of high-quality data, particularly in developing regions. The lack of consistent, up-to-date data for key variables
such as soil moisture, rainfall intensity, and river discharge can result in model inaccuracies (Muthukumar et al., 2020).
Additionally, the integration of real-time data into flood risk models remains a challenge, as does quantifying uncertainty in
model predictions (García et al., 2022).

IJMRA, Volume 7 Issue (Month) 2024 www.ijmra.in Page 2

Cite as :Paper Title; Vol. …|Issue ….|Pg:102-107 [Pick the date]
Future research should focus on addressing these challenges by improving data quality, enhancing model calibration, and
integrating real-time monitoring systems. Moreover, there is a need for more studies that integrate socio-economic factors into
flood susceptibility mapping, as these can significantly influence the impacts of flooding (Tao et al., 2020). As technology
continues to advance, the combination of machine learning, remote sensing, and real-time data holds great potential for
improving flood risk assessment and management.

III. METHODOLOGIES

Flood susceptibility mapping is an intricate process that involves the use of various methodologies to predict flood-prone areas.
The integration of Geographic Information Systems (GIS) and remote sensing technologies enhances the ability to analyze
spatial and temporal data. Below are the key methodologies used in the field, based on recent research.

A. Geospatial Data Collection

Remote sensing plays a critical role in flood susceptibility mapping by providing high-resolution spatial data. Satellites and aerial
imagery are often used to obtain data such as land cover, elevation, and water bodies. LiDAR (Light Detection and Ranging)
and synthetic aperture radar (SAR) are frequently employed to measure topographic features and surface water distribution.
The Digital Elevation Model (DEM) is a common output used to analyze terrain features such as slope, aspect, and elevation
(Akinyemi et al., 2020).

GIS is then used to store, manipulate, and analyze these geospatial datasets. GIS software tools (e.g., ArcGIS, QGIS) allow
researchers to perform spatial analysis like buffer analysis, overlay analysis, and distance measurement to study flood
dynamics and susceptibility.

In the study by Nasir et al. (2023), remote sensing data from satellite images were used to develop flood susceptibility models for
the Swat River Basin by analyzing elevation and land-use patterns.

B. Flood Susceptibility Indices

Researchers commonly use statistical indices and weighting methods to assess flood susceptibility. Methods like Frequency
Ratio (FR), Weight of Evidence (WOE), and Weighted Overlay are employed to identify the relative importance of various
factors in contributing to flood risks. These indices are particularly useful when combining diverse datasets such as soil type,
distance to rivers, rainfall, and land use.

a) Frequency Ratio (FR) calculates the probability of flood occurrence by determining the ratio of flood-prone areas for
different spatial variables (e.g., slope, land cover, soil type).
b) Weight of Evidence (WOE) assigns weights to different spatial factors based on their contribution to flood risk. These
weights are often derived from historical flood data or expert knowledge (Zhang et al., 2021).

Akinyemi et al. (2020) used the Frequency Ratio method in conjunction with GIS to assess flood susceptibility in Lagos,
Nigeria, finding that proximity to rivers and elevation significantly influenced flood occurrence.

C. Machine Learning Approaches

Machine learning algorithms, such as Random Forest (RF), Support Vector Machines (SVM), and Artificial Neural
Networks (ANN), are increasingly integrated with GIS and remote sensing data for flood susceptibility mapping. These
algorithms can handle complex and non-linear relationships between flood risk factors and are particularly useful when dealing
with large, multi-dimensional datasets.

a) Random Forest (RF) is an ensemble learning method that constructs multiple decision trees and aggregates their
predictions to improve accuracy.
b) Support Vector Machines (SVM) are supervised learning models used to classify flood-prone areas based on a training
dataset (e.g., elevation, rainfall, and proximity to water bodies).

c) Artificial Neural Networks (ANN) simulate the way the human brain processes information and can model complex,
non-linear interactions between flood risk variables.

García et al. (2022) applied Random Forest and SVM algorithms in an urban flood susceptibility study, achieving higher
accuracy compared to traditional methods by incorporating remote sensing data and GIS-based variables.

IJMRA, Volume 7 Issue (Month) 2024 www.ijmra.in Page 3

Cite as :Paper Title; Vol. …|Issue ….|Pg:102-107 [Pick the date]
D. Hydrological and Climate Modeling

To predict flood risk in dynamic environments, hydrological models (e.g., SWAT, HEC-RAS) are often combined with GIS and
remote sensing data. These models simulate the movement of water within watersheds, considering factors such as precipitation,
runoff, and land cover.

a) SWAT (Soil and Water Assessment Tool) is used for simulating rainfall-runoff processes in catchments to predict
flooding based on climate data and land-use changes.
b) HEC-RAS (Hydrologic Engineering Centers-River Analysis System) models river hydraulics to estimate flood flow
and depth, which can be incorporated into GIS-based flood maps.

These hydrological models can be adapted to include climate change projections, helping researchers assess future flood risks
under various climatic scenarios (Zhang et al., 2021).

Ahmad et al. (2022) used SWAT to model flood risk in the Kosi River Basin and integrated remote sensing data with
hydrological parameters to assess flood susceptibility under future climate conditions.

E. Accuracy Assessment and Validation

Ensuring the accuracy of flood susceptibility models is essential for reliable flood risk management. Various methods are used for
model validation, including confusion matrices, ROC curves, and cross-validation techniques.

a) Confusion Matrix: This tool compares predicted flood-prone areas with actual observed flood data, evaluating model
performance by calculating sensitivity, specificity, and accuracy.
b) Receiver Operating Characteristic (ROC) Curve: The ROC curve is used to assess the discriminative ability of flood
susceptibility models, with Area Under Curve (AUC) as a key performance indicator.

c) Cross-validation: This technique involves partitioning the data into subsets for training and testing, ensuring that the
model generalizes well to new data.

Muthukumar et al. (2020) used a confusion matrix to evaluate the accuracy of GIS-based flood susceptibility maps in Chennai,
India, comparing model results with actual flood data.

1. This section outlines the research methodology employed to assess the flood susceptibility of Cauayan
City, focusing on the integration of Digital Elevation Model (DEM) data, GIS tools, and statistical tests.
The methodology is designed to ensure the robustness and reliability of flood risk models and to address any
potential biases in the study.

2. Research Design and Approach

The study adopts a quantitative research approach to analyze flood susceptibility by integrating remote
sensing data, particularly DEM, with GIS-based flood models. The study focuses on examining variables
that influence flood risks, such as elevation, land cover, distance to rivers, and slope.

3. Data Collection

a. DEM Data
The DEM data used for this study is sourced from USGS Earth Explorer (SRTM) and ASTER GDEM,
which provide global elevation data at a resolution of 30 meters. These datasets are vital in determining the
topography of the area, which influences flood dynamics by dictating water flow patterns and the
vulnerability of different areas (Zhang et al., 2021).

b. Other Data Sources

Additional spatial data (e.g., land cover, soil type, climate data) are collected through satellite imagery
and ground-based surveys, ensuring that flood susceptibility models consider various environmental
factors.

IJMRA, Volume 7 Issue (Month) 2024 www.ijmra.in Page 4

Cite as :Paper Title; Vol. …|Issue ….|Pg:102-107 [Pick the date]
4. Research Tools and Software

a. ArcGIS
The ArcGIS software is used to perform spatial analysis and map the flood susceptibility in Cauayan City.
ArcGIS tools, such as Raster Analysis and Spatial Analyst, are used for buffer analysis, overlay analysis,
and proximity analysis to identify flood-prone areas based on topographic, hydrological, and land-use data
(Akinyemi et al., 2020).

b. Computational Resources
The analysis is conducted using an Asus ZenBook Intel Core i7 laptop with Nvidia graphics, providing
the necessary computational power to handle complex spatial analysis and machine learning models. The
processing power ensures efficient handling of large datasets and enhances the accuracy of flood risk
modeling.

5. Research Methods

a. Flood Susceptibility Mapping Using GIS

GIS is utilized to create a flood susceptibility map by analyzing the influence of various environmental
factors such as elevation, land cover, and distance to water bodies (Zhang et al., 2021). The model is built
using a combination of Frequency Ratio (FR), Weighted Overlay, and Weight of Evidence (WOE)
approaches.

b. Machine Learning Models

To enhance the flood susceptibility map's accuracy, machine learning algorithms such as Random Forest
(RF) and Support Vector Machines (SVM) are applied to the spatial data, as outlined by García et al.
(2022). These models help classify flood-prone areas by learning from historical flood data and spatial
features.

6. Tests Conducted

a. Accuracy Assessment
The accuracy of the flood susceptibility maps is evaluated using a confusion matrix, which compares the
predicted flood-prone areas with historical flood data. Metrics such as sensitivity, specificity, and overall
accuracy are computed to assess model performance (Muthukumar et al., 2020).

b. Sensitivity Analysis
A sensitivity analysis is performed to assess how changes in the input variables, such as elevation or land
cover, impact the flood susceptibility map. This helps identify the most critical factors affecting flood risks
(Nasir et al., 2023).

7. Ethical Considerations

The study adheres to ethical guidelines in data usage and analysis. All data sourced from satellite imagery,
DEM, and GIS databases are publicly available and used with proper attribution. The research avoids the use
of personal data or confidential information that might compromise privacy. Additionally, the study aims
to contribute positively to the community by providing actionable insights for flood risk management.

8. Steps Taken to Eliminate Biases

a. Data Selection
Careful consideration is given to the selection of spatial data, ensuring that the datasets used are up-to-date,
accurate, and representative of the study area. Historical flood data is also cross-verified with ground-truth
data to reduce potential biases in model predictions (Williams & Shaw, 2019).
IJMRA, Volume 7 Issue (Month) 2024 www.ijmra.in Page 5
Cite as :Paper Title; Vol. …|Issue ….|Pg:102-107 [Pick the date]
b. Model Calibration
Machine learning models are calibrated using a combination of training and testing datasets to minimize
overfitting. Cross-validation techniques are used to assess the model’s generalizability and to reduce biases
due to random fluctuations in the data (Cai et al., 2021).

c. Validation with Multiple Datasets

The flood susceptibility map is validated by comparing model predictions with independent datasets and
historical flood records. This ensures that the model’s results are reliable and not influenced by local
anomalies or biases in the data (Ahmad et al., 2022).

9. Conclusion

The methodology used in this study integrates advanced GIS tools, remote sensing data, and machine
learning models to create a robust flood susceptibility map for Cauayan City. By adhering to ethical
considerations and taking steps to eliminate biases, the research aims to produce reliable and accurate flood
risk assessments that can inform urban planning and flood management strategies.

RESULTS AND DISCUSSIONS

CONCLUSION

ACKNOWLEDGMENT

I would like to acknowledge Michael Shell and other contributors for developing and maintaining the IJMRA
LaTeX style files which have been used in the preparation of this template. Isabela State University Cauayan Campus my
Family and most of all our all-mighty God.

REFERENCES
1. Akinyemi, A. S., et al. (2020). GIS-based flood susceptibility mapping using remote sensing data in Lagos, Nigeria.
Journal of Environmental Management, 241, 123-130.
2. García, L., et al. (2022). Application of machine learning algorithms for flood susceptibility mapping: A case study in
urban areas. Remote Sensing Applications: Society and Environment, 25, 100-110.

3. Muthukumar, M., et al. (2020). Evaluation of flood susceptibility mapping using GIS and remote sensing data: A case
study in Chennai, India. Journal of Flood Risk Management, 15(4), 1-13.

4. Zhang, X., et al. (2021). Integration of GIS, remote sensing, and machine learning for flood susceptibility mapping.
Environmental Modeling & Software, 134, 104850.

5. Nasir, M. M., et al. (2023). Remote sensing and GIS-based flood risk assessment in the Swat River Basin, Pakistan.
Natural Hazards, 112(3), 1457-1475.

6. Williams, S., & Shaw, G. (2019). The role of GIS in flood risk assessment. Journal of Flood Risk Management, 12(1),
16-24.

7. Cai, X., et al. (2021). Application of deep learning algorithms for flood susceptibility mapping in the context of climate
change. Journal of Hydrology, 588, 125-137.

V. CONCLUSIONS
The version of this template is V2. Most of the formatting instructions in this document have been compiled by Causal
Productions from the IJMRA LaTeX style files. Causal Productions offers both A4 templates and US Letter templates for LaTeX
and Microsoft Word. The LaTeX templates depend on the official IJMRA tran.cls and IJMRA tran.bst files, whereas the Microsoft
IJMRA, Volume 7 Issue (Month) 2024 www.ijmra.in Page 6
Cite as :Paper Title; Vol. …|Issue ….|Pg:102-107 [Pick the date]
Word templates are self-contained. Causal Productions has used its best efforts to ensure that the templates have the same
appearance.
Causal Productions permits the distribution and revision of these templates on the condition that Causal Productions is
credited in the revised template as follows: “original version of this template was provided by courtesy of Causal Productions
(www.causalproductions.com)”.

IJMRA, Volume 7 Issue (Month) 2024 www.ijmra.in Page 7