Surigao del Norte National High School

Peñaranda St., Surigao City

Optimizing Heat Index Prediction: A Forecasting Through ARIMA

and Python-Based Approach for Climate Adaptation

A Capstone Project Mathematics and Computational Science

Category

Presented to the Faculty of Surigao del Norte National

High School Senior High School Peñaranda

St., Surigao City

In Partial Fulfillment of the Requirements for Senior High School

Capstone Project Science, Technology,

Engineering, and Mathematics

Mae Ann J. Cabug-os

Alexis D. Gio

Irene Jade D. Nono

Charise V. Orjela

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

March 2025

ABSTRACT

The increasing frequency of extreme heat events due to climate

change necessitates accurate heat index prediction for climate

adaptation. This study optimizes heat index forecasting using a

Python-based AutoRegressive Integrated Moving Average (ARIMA)

model. Traditional methods often fail to capture climate variability,

highlighting the need for improved predictive techniques.

The study integrates temperature, humidity, wind speed,

atmospheric pressure, and precipitation from weather APIs and

satellite data to enhance accuracy. Meteorological data were

analyzed using ARIMA, with its predictive performance compared to

conventional models. Results indicate that humidity strongly

correlates with heat index fluctuations, while wind speed plays a

crucial role in heat stress mitigation. ARIMA demonstrated superior

accuracy in short-term predictions, making it valuable for public

safety measures.

This research underscores the importance of incorporating

multiple meteorological variables in heat index forecasting. It

provides actionable insights for weather agencies, policymakers, and

communities, emphasizing real-time forecasting and hybrid modeling

approaches to enhance climate resilience and mitigate extreme heat

risks.

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

Key terms: Heat Index Prediction, ARIMA, Climate Adaptation,

Regression Model, Weather API, Satellite Data

APPROVAL SHEET

OPTIMIZNG HEAT INDEX PREDICTION: A FORECASTING THROUGH

PYTHON-BASED APPROACH FOR CLIMATE ADAPTATION

Mae Ann J. Cabug-os, Alexis D. Gio, Irene Jade D. Nono, and

Charise V. Orjela, 2025
In fulfillment of the requirements for the Science, Technology,

Engineering, and Mathematics (STEM) Strand, this research study has

been reviewed and is recommended for approval in Oral Examination.

ROCHELLE R. YLARAN
Research Teacher
-------------------------------------------------------------------------------------------
-----
APPROVED by the Committee for Oral Examination with a grade of PASSED.

________________________________

Chairman

___________________________ ______________________________

Member Member

-------------------------------------------------------------------------------------------

------

CERTIFIED that this Capstone Project is ACCEPTED in fulfillment of

the requirements for the Science, Technology, Engineering, and

Mathematics (STEM) Strand, this February 2025.

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

JULIET C. TIUJONGCO

SHS-School Principal

DINDO P. GIPALA

Secondary School Principal IV

ACKNOWLEDGEMENT
First and foremost, we, the researchers would like to express

our deepest gratitude to the Almighty God for granting us the

wisdom, strength, and perseverance to accomplish this research. His

guidance has been the foundation of this study, providing us with the

determination to overcome challenges along the way.

We would also like to extend our sincere appreciation to the

Philippine Atmospheric, Geophysical and Astronomical

Services Administration (PAGASA) Surigao station for

acknowledging our research and providing valuable insights, as well

as for granting access to essential climate data.

We are profoundly grateful to our research advisers, Ma’am

Rochelle Ylaran and Sir Rednil Labi, for their continuous support,

guidance, and expertise throughout the research process. Their

valuable feedback has been instrumental in refining this study.

Special thanks are given to our school principals, Mr. Dindo P.

Gipala, Mrs. Juliet C. Tiujunco, and Mrs. Lucyl L. Morales, for

their encouragement and for fostering an academic environment

conducive to research.

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

The researchers also wish to express their deep appreciation to

Mr. Vel Zoed Sungahid for his invaluable assistance in coding,

which significantly contributed to the development of the forecasting

model used in this study.

Lastly, we would like to extend our heartfelt gratitude to our

beloved parents, whose unwavering love, patience, and support

have been our greatest source of motivation. Their encouragement

has been crucial in completing this research.

Without the collective efforts of these individuals and

institutions, this study would not have been possible.

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

TABLE OF CONTENTS

Title Page ----------------------------------------------------

Abstract ----------------------------------------------------

ii

Approval Sheet ----------------------------------------------------

iii

Acknowledgement ----------------------------------------------------

iv

Table of Contents ----------------------------------------------------

vi

List of Tables ----------------------------------------------------

viii

List of Figures ----------------------------------------------------

ix

1 The Problem and its Background

----------------------------- 1

Introduction ----------------------------- 1

Review of Literature and Studies --------------------------- 3

Theoretical Framework ---------------------------- 20

Conceptual Framework -------------------- 21

Statement of the Problem -----------------------------

22

Objectives of the Study ---------------------------- 24

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

Hypothesis ----------------------------- 24

Significance of the Study ----------------------------- 25

Scope and Limitation of the Study ----------------------------- 27

Definition of Terms ---------------------------- 28

2 Research Methodology

Research Design --------------------------- 30

Research Locale --------------------------- 32

Materials ----------------------------- 36

Data Gathering Procedure -------------------------- 34

Data Analysis -------------------------- 38

3 Results and Discussion

4 Summary of Findings, Conclusion, and Recommendation

Summary of Findings ----------------------------- 53

Conclusions --------------------------- 56

Recommendations -----------------------------

57

References

Appendices

A. Letter of Acknowledgement

B .Documentation

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

C. Coding

Curriculum Vitae

LIST OF FIGURES

Table Title Page

1 Schematic Diagram of the Study ………………………………. 21

2 Location Map of the Study …………………………….…… 32

3 Flowchart of the Study …………………………….………… 37

PAGE \* MERGEFORMAT i
Surigao del Norte National High School
Peñaranda St., Surigao City

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

CHAPTER 1

THE PROBLEM AND REVIEW OF LITERATURE

This chapter presents the introduction, related literature, conceptual

framework, statement of the problem, hypothesis, significance of the

study, scope and limitation of the study, and definition of terms.

INTRODUCTION

The increasing frequency and intensity of extreme heat events due

to climate change have made accurate heat index prediction a crucial

component of climate adaptation. Prolonged exposure to extreme heat

has been linked to various health risks, including heat exhaustion, heat

stroke, and cardiovascular diseases. Moreover, rising temperatures impact

economic activities, productivity, and overall human well-being. Thus, the

ability to forecast heat index levels accurately is essential for public safety

and climate resilience.

Traditional forecasting techniques, while useful, often fail to capture

the complexity of meteorological data and climate variability. Many

existing methods rely on historical climate patterns and simplistic

modeling approaches that may not fully account for dynamic changes in

temperature and humidity. Regression models have been widely used in

climate science to establish relationships between temperature, humidity,

and other meteorological variables for heat index prediction

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

(Adityasrambhad et al., 2022). However, improving their accuracy requires

incorporating modern computational tools and statistical techniques.

Python, with its extensive ecosystem of statistical and data science

libraries, has emerged as a powerful tool for climate modeling. Libraries

such as NumPy and Pandas facilitate efficient data preprocessing, while

SciPy and Stats models enable advanced statistical computations.

Additionally, visualization tools like Matplotlib and Seaborn support

researchers in interpreting climate trends. By leveraging these tools, this

study aims to optimize heat index prediction through a robust Python-

based regression model.

Despite their advantages, regression models face challenges such as

data inconsistencies, missing values, and overfitting. Ensuring reliability

requires robust validation techniques, such as cross-validation and

residual analysis, to improve predictive accuracy. This research explores

the role of regression models in optimizing heat index prediction for

climate adaptation, with a specific focus on Surigao del Norte.

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

REVIEW OF LITERATURE

The foundation of this study is supported by the reviewed literature

of which were organized thematically.

Comparison between regression models and machine learning for

heat index prediction accuracy, strengths, and weaknesses

The comparison between regression models and machine learning

(ML) techniques for heat index prediction is a crucial area of study in

meteorology and data science. Both approaches offer distinct advantages

and limitations when modeling the relationship between meteorological

variables and the resultant heat index, an essential measure for public

health and safety in extreme weather conditions. With climate change

exacerbating the occurrence of heatwaves, understanding the strengths

and weaknesses of these methods becomes increasingly important for

accurate heat index forecasting.

Regression models, such as linear and polynomial regression, are

often favored for their simplicity, interpretability, and statistical rigor.

These models allow researchers to identify clear relationships between

variables, making them particularly valuable in scenarios requiring a

deeper understanding of how each independent variable influences the

dependent one. However, regression models struggle to capture complex,

non-linear interactions between variables without the introduction of

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

higher-degree terms, limiting their predictive accuracy in more intricate

scenarios (Singh, 2024).

In contrast, machine learning models are adept at handling large

datasets with numerous complex features, automatically identifying

patterns that may not be immediately apparent in traditional regression

models. They have the advantage of adaptability, especially in dynamic

environments where relationships evolve over time. However, this

strength also presents challenges in terms of interpretability, as many

machine learning algorithms function as "black boxes," raising concerns

about the transparency of predictions, especially in critical areas such as

public health ("Comparing ML and Statistical Models: Effectiveness and

Performance," 2020).

Despite these differences, there is increasing interest in hybrid

approaches that combine the strengths of both regression and machine

learning methods. Such hybrid models aim to improve prediction accuracy

while maintaining the interpretability that stakeholders require, making

them a promising approach for more robust heat index forecasting

("Choosing the Right Machine Learning Model: A Comprehensive Guide,"

2020; Sourjah & Pemarathne, 2022).

Regression Models

Regression models are fundamental in statistical modeling and serve

as a reliable method for identifying relationships between a dependent

variable and one or more independent variables. Particularly in the case of

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

heat index prediction, regression models help establish linear or nonlinear

relationships between meteorological features such as temperature,

humidity, and wind speed.

Polynomial Regression

Polynomial regression, an extension of linear regression, allows for

the modeling of non-linear relationships by introducing higher-degree

terms. This method is particularly useful when the relationship between

the dependent and independent variables cannot be captured accurately

with a linear model alone. By adding terms such as x2x^2x2, x3x^3x3,

etc., polynomial regression offers greater flexibility and can provide more

accurate predictions for complex systems like heat index forecasting

(Singh, 2024)

Assumptions and Limitations

Polynomial regression shares many of the assumptions of traditional

linear regression, such as linearity in terms of the unknown parameters,

independence of observations, and homoscedasticity of errors. Violations

of these assumptions can lead to biased or inefficient estimates,

highlighting the need for careful diagnostic checks. Additionally, while

polynomial regression can model nonlinear relationships, it is prone to

overfitting, especially when higher-degree polynomials are used (Singh,

2024).

Strengths of Regression Models

One key strength of regression models is their interpretability. The

coefficients of a regression model provide clear insights into how each

variable affects the outcome. This transparency is essential in fields like

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

public health, where understanding the relationship between

environmental factors and health outcomes is vital.

Furthermore, regression models are computationally less

demanding compared to machine learning models, making them easier to

implement in situations where computational resources are limited

("Comparing ML and Statistical Models: Effectiveness and Performance,"

2020).

Machine Learning Models

Machine learning models have revolutionized many predictive fields,

including meteorology. These models can handle large, complex datasets

and are particularly adept at capturing non-linear relationships between

multiple variables. In heat index prediction, machine learning techniques

such as decision trees, support vector machines (SVMs), and neural

networks have shown great promise in providing accurate forecasts.

Supervised Learning and Heat Index Prediction

In supervised learning, machine learning algorithms learn from

labeled data to predict outcomes based on input features. In the context

of heat index prediction, this could involve training a model on historical

weather data to predict the heat index under different conditions. Decision

trees, for example, can be used to create simple decision rules based on

the most relevant meteorological features, while SVMs can find

hyperplanes that optimally separate data points for classification (Sourjah

& Pemarathne, 2022).

Advantages of Machine Learning Models

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

One of the main advantages of machine learning is its ability to

process large, high-dimensional datasets. These models can automatically

learn features that are most important for predicting the heat index,

removing the need for manual feature selection or transformation.

Moreover, ML models can adapt to changes in data over time, which is

particularly valuable in environments where weather patterns are dynamic

("Choosing the Right Machine Learning Model: A Comprehensive Guide,"

2020).

However, machine learning models come with their own set of

challenges. One of the primary concerns is their lack of interpretability.

Unlike regression models, where the relationships between variables are

explicitly defined, machine learning models often operate as "black

boxes," making it difficult to understand why a model made a particular

prediction ("Pros and Cons of Various Machine Learning Models: A

Comparison," 2020). This can be problematic in sectors like public health,

where transparency is crucial for decision-making.

Comparison of Regression Models and Machine Learning

Performance and Accuracy

Both regression models and machine learning techniques have their

strengths, and their performance depends on the nature of the data.

Regression models are easier to interpret and require fewer computational

resources, but they may struggle to accurately predict heat index values

when the relationship between variables is complex. In contrast, machine

learning models excel at capturing these complex, non-linear relationships

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

but require larger datasets and more computational power ("Comparing

ML and Statistical Models: Effectiveness and Performance," 2020).

Hybrid Approaches

Given the strengths and weaknesses of both regression and machine

learning models, hybrid approaches that combine both methods are

gaining traction. These models aim to leverage the interpretability of

regression while incorporating the adaptability and accuracy of machine

learning. This approach may offer the best of both worlds, making

predictions more accurate while still providing insights into the decision-

making process ("Choosing the Right Machine Learning Model: A

Comprehensive Guide," 2020; Sourjah & Pemarathne, 2022).

Why Regression Models Are Sometimes Preferred Over Machine

Learning in Climate Research

Simplicity and Interpretability

Regression models are often favored in climate research due to their

simplicity and interpretability. Unlike machine learning (ML) models, which

can function as "black boxes," regression techniques provide clear

relationships between input variables and outcomes. According to

"Assumptions in Regression: Why, What, and How" (2020), regression

models, such as linear regression, allow researchers to easily understand

and communicate the effects of climate variables. This interpretability is

crucial in policymaking and scientific communication, as stakeholders

require transparent models to make informed decisions.

Statistical Rigor and Hypothesis Testing

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

Regression models provide a statistically rigorous framework for

hypothesis testing, confidence interval estimation, and assessing

relationships between climate variables. "An Introduction to the Statistical

Modelling of Climate Change: 1850-2020" (2021) highlights how

regression models facilitate hypothesis testing, ensuring that findings are

statistically significant rather than artifacts of noisy data. In contrast,

many ML models lack inherent statistical validation, making their

predictions more difficult to interpret and validate.

Handling Non-Stationarity in Climate Data

Climate data often exhibit non-stationary characteristics, meaning

that statistical properties change over time. Regression models,

particularly time series regression techniques, are designed to handle non-

stationarity. "Climate Change Forecasting Using Machine Learning

Algorithms" (2021) notes that regression methods can incorporate

autoregressive terms and differencing techniques to address changing

climate trends effectively. Machine learning models, on the other hand,

require extensive preprocessing to handle nonstationarity, often leading to

increased complexity without necessarily improving predictive

performance.

Flexibility in Model Design

Regression models offer flexibility in their design, allowing

researchers to incorporate polynomial regression, Bayesian regression,

and generalized linear models based on specific research needs. "A

Bayesian Framework for Emergent Constraints: Case Studies of Climate"

(2023) explains how Bayesian regression methods can integrate prior

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

knowledge and uncertainty quantification, making them more adaptable

for climate projections.

This flexibility ensures that regression models remain relevant in

various climate research applications.

Efficient Use of Limited Data

In many climate studies, the availability of high-quality

observational data is limited. Regression models perform well even with

smaller datasets, whereas machine learning models often require large

datasets to generalize effectively. "Evaluating and Calibrating Uncertainty

Prediction in Regression Tasks" (2022) states that regression models can

provide meaningful insights with fewer observations, making them

particularly useful in historical climate data analysis.

Limitations of Machine Learning Models in Climate Research

While ML models are increasingly used in climate studies, they

present several challenges. "Advancing Interpretability of Machine-

Learning Prediction Models" (2023) discusses how the opacity of deep

learning models makes it difficult for researchers to validate results and

ensure alignment with established climate science principles. Additionally,

"The Mythos of Model Interpretability" (2020) points out that complex ML

algorithms can capture spurious correlations, leading to unreliable

predictions.

Ethical and Reliability Concerns

Machine learning models are susceptible to biases due to training

data limitations, which can result in misleading climate predictions.

"Interpretable Machine Learning for Weather and Climate" (2022) warns

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

that biased datasets can produce inconsistent results, especially for

underrepresented regions. Moreover, model drift—where predictive

accuracy degrades over time—is a known issue in ML, necessitating

continuous retraining and validation.

Physical Consistency and Model Validation

Unlike regression models, which rely on well-established statistical

principles, many ML models do not inherently enforce physical constraints.

"GMD - Machine Learning for Numerical Weather and Climate Modelling"

(2023) emphasizes the need to integrate physics-based constraints into

ML models to improve their reliability in climate applications. Regression

models, by contrast, inherently maintain physical consistency due to their

structured approach.

Hybrid Approaches and Future Directions

Despite their differences, regression models and machine learning

techniques are increasingly being integrated to leverage their respective

strengths. "Assessing the Role of Machine Learning in Climate Research"

(2024) suggests that hybrid models combining regression and ML can

enhance predictive accuracy while retaining interpretability. As research

evolves, the emphasis on explainable AI and physics-informed machine

learning is expected to bridge the gap between traditional statistical

methods and modern computational techniques.

Recent Advancements in Real-Time Climate Forecasting and Their

Impact on Heat Index Predictions

Advancements in real-time climate forecasting have revolutionized

the ability to predict extreme heat events and assess their impact on the

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

heat index. These improvements are critical as climate change continues

to intensify the frequency and severity of extreme weather patterns. "The

Climate Technology Progress Report 2023" (2023) highlights how the

integration of machine learning (ML) and artificial intelligence (AI) in

climate models has enhanced the precision of forecasting systems,

allowing for more accurate and timely predictions.

Evolution of Weather Forecasting Techniques

Historically, weather forecasting relied on numerical weather

prediction (NWP) models based on physics-driven equations. However, the

emergence of AI has allowed for data-driven forecasting techniques that

significantly improve accuracy and computational efficiency. "Ten New

Insights in Climate Science 2023" (2023) discusses how AI models can

process vast datasets and detect complex atmospheric patterns that

traditional models may overlook. This evolution has been crucial in

addressing the increasing frequency of extreme weather events and their

impact on public health.

The Role of Machine Learning in Modern Forecasting

Recent studies have demonstrated the effectiveness of ML in

improving climate prediction accuracy. "Building a Weather Prediction

Model with Machine Learning" (2023) details how ML algorithms, such as

deep learning models, analyze historical climate data to identify patterns

and predict future weather conditions with greater precision. Similarly,

"GMD - Machine Learning for Numerical Weather and Climate Modelling"

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

(2023) explores the integration of ML techniques with traditional climate

models, leading to hybrid approaches that optimize short-term and long-

term predictions.

Impact on Heat Index Predictions

One of the most significant applications of real-time forecasting

advancements is in heat index predictions. "Climate Change: Improved

Prediction of Heatwaves Thanks to AI" (2023) explains how AI-driven

models predict extreme heat events by incorporating environmental

variables such as temperature, humidity, and soil moisture. These models

enable authorities to implement proactive measures, such as heat

warnings and public health interventions, to mitigate the impact of

extreme temperatures.

Challenges and Limitations

Despite these advancements, challenges remain in ensuring data

accuracy and accessibility. "Energy End-Use Data Collection Methodologies

and the Emerging Role of Digital Technologies" (2023) highlights the

difficulties in managing and standardizing climate data across different

regions. Moreover, "Interpretable Machine Learning for Weather and

Climate" (2022) raises concerns about the transparency of AI-driven

predictions, emphasizing the need for explainable models to enhance trust

and usability in decision-making processes.

Future Directions

To address these challenges, researchers are focusing on hybrid

forecasting models that integrate physics-based and AI-driven approaches.

"Assessing the Role of Machine Learning in Climate Research" (2024)

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

discusses the development of hybrid models that leverage the strengths of

both methodologies to improve prediction reliability.

Additionally, "How Can Artificial Intelligence Help Improve Climate

Forecasting and Risk Information?" (2023) explores new AI applications

that enhance real-time risk assessment for climate adaptation strategies.

The rapid advancements in real-time climate forecasting have significantly

improved heat index predictions, enabling better preparedness for

extreme heat events.

While challenges remain in data management and model

interpretability, ongoing research into hybrid models and AI-driven

solutions holds promise for further refining predictive capabilities. As

climate change continues to pose global risks, these technological

innovations will play a crucial role in enhancing resilience and public

safety.

Common Biases in Climate Datasets and Strategies to Address

Missing Data and Sensor Inaccuracies

Climate datasets play a crucial role in environmental research,

policy-making, and forecasting. However, various biases can compromise

the reliability of these datasets, affecting climate predictions and decision-

making processes. "Know Your Climate Bias" (2020) highlights that biases

in climate datasets stem from inconsistencies in data collection methods,

model assumptions, and sensor inaccuracies. Addressing these issues is

vital for ensuring the accuracy and usability of climate data.

Common Biases in Climate Datasets

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

Climate datasets can exhibit biases due to methodological

discrepancies, data inconsistencies, and model parameterization errors.

"Learning to Correct Climate Projection Biases" (2021) discusses how

biases in climate models can arise from over-reliance on historical data

that may not fully capture emerging climate trends. Additionally,

"Appendix A: Observational Datasets Used in Climate Studies" (2020)

explains that systematic biases can be introduced when datasets from

different sources are merged without proper calibration.

Recognizing and Addressing Biases

One of the first steps in addressing biases is recognizing their

sources. "Algorithmic Bias Detection and Mitigation" (2022) emphasizes

that biases in climate data can be similar to those found in other fields,

where certain data points are overrepresented while others are

underrepresented. Recognizing these biases allows researchers to develop

correction strategies, ensuring more balanced datasets.

Strategies to Address Missing Data

Missing data is a persistent issue in climate datasets, leading to

gaps that hinder comprehensive analysis. "GitHub - Missing Datasets"

(2021) describes missing climate data as a significant challenge that

requires a combination of statistical techniques and collaborative

approaches for resolution. "Closing Climate and Disaster Data Gaps: New

Challenges, New Thinking" (2022) suggests that international collaboration

can help address missing data by sharing information across institutions

and regions.

Methods for Handling Missing Data

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

Various methods exist for handling missing data, including :

Imputation Techniques: "METEO 810: Weather and Climate Data Sets"

(2023) explains that interpolation and statistical imputation methods can

estimate missing values based on available data.

Community-Based Data Collection: Engaging local communities in climate

data collection can fill gaps in observational records. Standardized Data

Collection Protocols: Ensuring consistency in data collection methods can

reduce missing data and enhance dataset reliability.

Strategies to Address Sensor Inaccuracies

Sensor inaccuracies can introduce errors in climate datasets,

affecting their precision and reliability. "Sensor Calibration Techniques for

Improved Reliability and Precision" (2023) highlights that calibration is

essential for ensuring data accuracy. "Essential Sensor Calibration

Techniques for Accurate Measurements" (2021) recommends in-situ

calibration, field validation, and routine maintenance as effective methods

for improving sensor reliability.

Standardization and Interoperability

To minimize biases and inaccuracies, climate researchers advocate

for standardization across datasets. "Specifying Calibration of

Environmental Sensors" (2023) emphasizes that establishing universal

calibration standards ensures comparability across different climate data

sources. "Sensor Calibration and Characterization to Meet Climate

Monitoring Requirements" (2022) further supports this by outlining best

practices for maintaining sensor accuracy.

Advanced Methodologies and Future Directions

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

Emerging methodologies such as AI-driven bias correction and

uncertainty quantification are being explored to improve climate data

accuracy. "Improving Climate Data Records with FIDUCEO" (2023)

describes how machine learning techniques can correct systemic biases

and enhance dataset reliability. Additionally, "Fidelity and Uncertainty in

Climate Data Records" (2024) underscores the importance of integrating

uncertainty quantification techniques to improve confidence in climate

projections. Addressing biases in climate datasets is essential for

improving the reliability of climate predictions. By implementing

standardized calibration techniques, utilizing statistical methods for

missing data, and fostering collaborative efforts, researchers can enhance

data accuracy. As technological advancements continue, integrating AI-

based methodologies and community-driven approaches will further

strengthen climate data integrity.

The Socioeconomic Impact of Rising Heat Index Levels on Outdoor

Workers, Farmers, and Student

The rising heat index, driven by climate change, poses

significant socioeconomic challenges for outdoor workers, farmers, and

students. Increasing temperatures and humidity levels are leading to

adverse health effects, reduced productivity, and disruptions in education.

"Heat Hazards and Migrant Rights" (2023) highlights how extreme heat

disproportionately affects marginalized communities, exacerbating

existing economic inequalities.

Impact on Outdoor Workers

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

Outdoor workers, particularly those in construction, agriculture,

and transportation, are among the most vulnerable to rising heat index

levels. "Industry Partnerships Create Policy Tools to Protect Outdoor

Laborers in Extreme Heat" (2023) discusses how excessive heat exposure

leads to increased cases of heat-related illnesses, lost work hours, and

reduced earnings. Furthermore, "Turner Studying Effects of Rising

Temperatures on Construction Workers" (2023) reveals that prolonged

exposure to high temperatures has resulted in lower productivity and

higher accident rates in labor-intensive industries.

Health Risks and Socioeconomic Consequence

Workers experiencing heat stress often suffer from dehydration,

heat exhaustion, and in extreme cases, heat stroke. "Rising Heat Risks for

Workers" (2023) reports that heat-related illnesses lead to a significant

number of workplace injuries and fatalities each year. Additionally,

"Feeling the Heat – The Effects of Higher Temperatures on People, Crops,

and Livelihoods" (2023) emphasizes that prolonged exposure to high

temperatures increases long-term health risks, including cardiovascular

diseases, affecting workers' ability to sustain their livelihoods.

Impact on Farmers

Farmers are particularly vulnerable to rising temperatures, as

excessive heat reduces crop yields and threatens livestock health. "The

Impact of High Temperatures on Livestock" (2023) highlights that heat

stress in animals leads to decreased productivity and increased mortality

rates. Furthermore, "Quantifying the Impact of Future Extreme Heat on the

Outdoor Work Sector in the United States" (2023) explains how extreme

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

temperatures lead to lower agricultural output and increased operational

costs.

Economic and Food Security Challenges

The agricultural sector is witnessing a decline in income due to climate-

induced losses. "2022 Young Farmers Climate and Water Policy

Recommendations" (2022) notes that extreme heat affects soil moisture

retention, leading to water shortages and higher irrigation costs. Additionally, "11

Ways Farmers Are Adapting to the Unpredictability of Climate Change" (2023)

highlights adaptive strategies such as drought-resistant crops and precision

agriculture, which can mitigate some of the negative impacts but require

significant investment.

Impact on Students

Students are also experiencing the effects of rising heat index levels,

particularly in regions where schools lack adequate cooling systems. "The

School Year Is Getting Hotter: How Does Heat Affect

Student Learning and Well-being?" (2022) discusses the negative

correlation between high classroom temperatures and student

performance. Research indicates that students exposed to excessive heat

during exams score significantly lower compared to those in cooler

conditions.

Educational Disruptions and Health Concerns

Heat waves often lead to school closures, disrupting learning

schedules. "Heatwaves Pose Health Risks, Disrupt Students’ Learning"

(2023) explains how high temperatures can impair cognitive function,

increase fatigue, and reduce concentration levels. Additionally, students in

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

low-income communities are disproportionately affected, as many schools

in these areas lack air conditioning and proper ventilation systems.

Strategies for Mitigation and Adaptation

To address these challenges, various mitigation strategies are being

proposed: For Outdoor Workers: Implementation of heat safety

regulations, increased access to hydration stations, and adjusted work

schedules, as suggested by "Outdoor Workers Face Increased Vulnerability

from Weather" (2023).

For Farmers: Adoption of climate-smart agricultural practices,

investment in resilient crop varieties, and improved irrigation techniques,

as highlighted in "Protecting Agricultural Workers in a Changing Climate"

(2023).

For Students: Upgrading school infrastructure with better cooling

systems, scheduling classes during cooler hours, and increasing public

awareness of heat-related health risks, as emphasized in "The School Year

Is Getting Hotter" (2022)

THEORETICAL FRAMEWORK

This study is anchored on Dahl et al. (2019), who emphasize that the

heat index (HI) is a crucial metric calculated using meteorological data

such as temperature and humidity. Regression analysis is widely used to

describe variations in heat index values, enabling researchers to develop

predictive models that help communities adapt to extreme heat

conditions.

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

According to Kjellstrom et al. (2019), heat stress is influenced by

multiple environmental factors, including temperature, humidity, radiant

heat, and air velocity. Understanding these variables allows for the

development of predictive models that provide early warnings for

heatwaves. Additionally, modern climate forecasting frameworks

emphasize the need for statistical validation techniques to ensure the

reliability of predictive models.

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

CONCEPTUAL FRAMEWORK

Input Process Output

 Historical heat
index data  Feature
Selection and  Math modeling
Regression by python-
 Climatic Model based
factors Development
(temperature 
(Python-
and humidity) Driven)

Figure 1. Schematic Diagram

"The figure above illustrates the input, process, and output of our

study in predicting the heat index using a Python-based regression model.

It shows that the prediction accuracy metric results from the developed

regression model are the output from the historical data.". It shows that

the prediction accuracy metric results from the developed regression

model is the output from the historical data.

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

STATEMENT OF THE PROBLEM

This study aimed to optimize heat index prediction using a

Python-based ARIMA approach to support climate adaptation efforts.

Through data collection and analysis, the study successfully identified

key meteorological factors influencing heat index predictions and

evaluated the effectiveness of regression models compared to

existing forecasting techniques.

Specifically, the study addressed the following research questions and

1. How can the ARIMA-based model optimize heat index prediction

compared to traditional forecasting methods?

2. How do meteorological variables (humidity, wind speed,

atmospheric pressure, and precipitation) contribute to improving

heat index forecasting accuracy?

By addressing these research questions, the study contributes to the

advancement of heat index prediction models, ultimately supporting

climate adaptation efforts in Surigao del Norte and similar climate-

sensitive regions.

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

OBJECTIVES OF THE STUDY

The objective of this study is to develop a regression model for

optimizing heat index prediction by integrating multiple meteorological

variables.

The objectives of the study are the following:

 To identify relevant meteorological variables e.g. temperature,

relative humidity, wind speed and, solar radiation.

 To gather historical meteorological data for the variables identified.

 To enhance predictive accuracy by incorporating various

meteorological variables

 To tailor the model to specific geographic regions, recognizing that

local climate conditions can significantly affect heat index

calculations.

HYPOTHESIS

At 0.05 level of significance, it is hypothesized that:

H₀: There is no significant effect of using a regression model in optimizing

heat index prediction for climate adaptation.

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

SIGNIFICANCE OF THE STUDY

The findings of the study may give benefit/s to the following:

Weather Forecast Agency. This study offers an improved

regression model for heat index prediction, aiding weather agencies

in issuing timely and accurate warnings to enhance public safety

during extreme heat.

School administrators. The study provides valuable tools for

improving policies and infrastructure. It enables better decision-

making, such as implementing heat safety guidelines, improving

ventilation systems, and developing long-term strategies to manage

heat-related risks.

Communities. This research offers practical benefits by providing

timely heat index forecasts. This allows people to plan their activities,

protect their health, and reduce the risks associated with extreme

heat.

Students. It ensures their safety and well-being by helping schools

plan better for hot weather. By providing accurate heat index

predictions, schools can adjust outdoor activities, enhance cooling

measures, and teach students about the effects of climate change

and how to adapt responsibly.

Teachers. The study supports them in creating safe and effective

learning environments. With reliable forecasts, they can adapt lesson

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

plans, organize indoor activities during extreme heat, and educate

students about climate adaptation.

Future researchers. The study contributes to the field of climate

science by showcasing the potential of regression models in

improving heat index predictions. It serves as a foundation for further

studies aimed at enhancing climate adaptation strategies.

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

SCOPE AND LIMITATION OF THE STUDY

The study focuses on optimizing heat index prediction using a

Python-based regression model for Surigao del Norte National High School.

Data is primarily sourced from NOAA and PAGASA, emphasizing

temperature and humidity trends in Surigao City.

However, limitations exist:

 The accuracy of predictions decreases with distance from the study

area.

 The study does not account for localized microclimatic effects such

as urban heat islands.

 External variables such as wind speed and solar radiation are not

included in the initial model but may be explored in future research.

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

DEFINITION OF TERMS

The following terms are defined conceptually and operationally for

better understanding of the readers with regards to the content and

context of this study.

APPARENT. It refers on how hot or cold feels to a human, which may

differ from the actual air temperature due to factors like humidity or

wind.

API. Stands for Application Programming Interface, which refers to a

set of rules and protocols that allow different software applications to

communicate and exchange data with each other; essentially, it's a

way for different programs to interact and share information in a

standardized manner.

ATMOSPHERIC MOISTURE. It refers to the amount of water vapor

present in the atmosphere. It plays a crucial role in weather patterns,

the water cycle, and the Earth's climate system.

CROSS-VALIDATION - A technique for testing the reliability of a

predictive model.

PHYTHON - A high-level programming language used for data

analysis and modeling.

DETRIMENTAL - It means causing harm or damage

HEAT INDEX - A measure of perceived temperature when relative

humidity is factored in.

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

METEOROLOGICAL VARIABLES. These are weather factors or

conditions that we can measure.

OPTIMIZE - means to make something as good or effective as

possible.

PREDICTION - means using past data to guess what will happen in

the future or estimate unknown values. It's like making an informed

guess based on patterns or trends.

REGRESSION MODEL- A mathematical model used to predict the

value of one variable based on another.

RELATIVE HUMIDITY - is the amount of water vapor present in the

air compared to the maximum amount the air can hold at a given

temperature.

CLIMATE ADAPTATION - Strategies implemented to minimize the

adverse effects of climate change on communities and ecosystems.

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

CHAPTER 2

METHODS

This chapter presents the research design, research environment,

validity and reliability, materials, data gathering procedure and data

analysis.

RESEARCH DESIGN

This study employs a time-series analysis approach to optimize heat

index prediction using a Python-based model.

Time-Series Analysis

Time-series analysis is used to analyze past heat index data and

forecast future values. This method examines trends, seasonal variations,

and anomalies in meteorological data over time. The study will employ

Autoregressive Integrated Moving Average (ARIMA) to improve the

accuracy of the predictions. These techniques help in capturing complex

temporal patterns and dependencies in heat index fluctuations.

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

RESEARCH ENVIRONMENT

Research study is conducted in Surigao del Norte National High

School located at Peñaranda St., Surigao City. It can be reached through

riding a tricycle, multicab or even motor bike.

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

(Source:https://www.google.com/maps/place/

Surigao+del+Norte+National+High+School/@9.785634,125.492889,15z/

data=!4m2!3m1!1s0x0:0x34f3b53acede8572?

sa=X&ved=2ahUKEwjMhMz_reyDAxUsmq8BHcWWC_8Q_BJ6BAgQEA)

Figure 2. Map of the Study

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

VALIDITY AND RELIABILITY

To ensure the accuracy and credibility of the study, data is

obtained from reliable meteorological sources, including:

 NOAA (National Oceanic and Atmospheric Administration) –

Provides historical and real-time global climate data.

 PAGASA (Philippine Atmospheric, Geophysical and Astronomical

Services Administration) – Supplies localized climate data

specific to the Philippines.

 Weather APIs (OpenWeather, Climate Data API, Stormglass API)

– Used for real-time temperature and humidity readings.

The study employs cross-validation techniques to test the

reliability of the proposed model. Additionally, outlier detection and

data preprocessing techniques such as Z-score normalization and

missing value imputation will be applied to maintain dataset integrity.

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

DATA GATHERING PROCEDURE

1. Collect Data: Gather temperature, humidity, wind speed, and

atmospheric pressure data from wunderground (Satelites), and

Weather APIs (GlassStorm API) for the past 3 months.

2. Organize Data: Store the collected data in a structured format

(CSV or database) for easy processing.

3. Check for Errors: Remove duplicate entries and fill in missing

values to ensure consistency.

4. Apply ARIMA Model: Use Python to analyze trends and forecast

heat index values.

5. Compare Predictions: Compare the ARIMA model's forecasts

with wunderground (Satelites), and Weather APIs (GlassStorm

API records to assess accuracy.

6. Data Collection: Historical temperature and humidity data will

be collected from wunderground (Satelites), and Weather APIs

(GlassStorm API) for the past 3 months to ensure a robust

dataset for time-series modeling.

7. Basic Data Cleaning: Remove duplicate entries, correct missing

values, and ensure consistency in temperature and humidity

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

readings. The data will be cleaned to handle missing values,

standardize units, and remove anomalies that could distort

predictions.

8. Model Development: Apply the ARIMA model to the collected

dataset to generate heat index predictions. Important

meteorological variables such as temperature, humidity, wind

speed, and atmospheric pressure will be considered in the

model.

9. Model Development:

 ARIMA Model: Used for time-series forecasting and

trend analysis. Results will be compared with

existing forecasting models from wunderground

(Satelites), and Weather APIs (GlassStorm API)

10. Basic Validation: Compare forecasted values with actual

recorded data to assess the model's effectiveness. Basic

validation will involve checking whether forecasted heat index

values align with actual observations over the study period.

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

Materials:

 Computer/Laptop – For data processing, model development, and

simulation.

 Python Programming Language – Used for data analysis and

modeling (with libraries such as Pandas, NumPy, SciPy, Matplotlib,

Seaborn, and TensorFlow for LSTM).

 Climate Datasets – Obtained from NOAA, PAGASA, and Weather APIs.

 Jupyter Notebook/Google Colab – For running Python scripts and

performing model training.

 Statistical Software (SPSS/R) – Used for additional statistical analysis

when comparing models.

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

Figure 3. Flow Chart of the General Procedures

PAGE \* MERGEFORMAT i
Surigao del Norte National High School
Peñaranda St., Surigao City

DATA ANALYSIS

Temp °F

Humidity %

PAGE \* MERGEFORMAT i
Surigao del Norte National High School
Peñaranda St., Surigao City

Heat Index °C

wunderground (Satelite)

PAGE \* MERGEFORMAT i
Surigao del Norte National High School
Peñaranda St., Surigao City

Temp °F

PAGE \* MERGEFORMAT i
Surigao del Norte National High School
Peñaranda St., Surigao City

Heat Index

GlassStorm (API)

PAGE \* MERGEFORMAT i
Surigao del Norte National High School
Peñaranda St., Surigao City

Temperature °F

Humidity %

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

Heat Index °C

Data analysis plays a crucial role in understanding and

predicting the heat index by examining key meteorological factors

such as temperature, humidity, wind speed, and atmospheric

pressure. In this study, data was collected from Wunderground

(Satellite) and GlassStorm (API), ensuring accuracy in forecasting.

The process involved cleaning and preprocessing the data by

handling missing values, removing outliers, and standardizing units.

The ARIMA (AutoRegressive Integrated Moving Average)

model was then applied to identify trends and patterns in historical

heat index data, allowing for more precise short-term predictions.

The results revealed a strong correlation between humidity

and heat index, confirming that even moderate temperature

increases can cause extreme heat conditions if humidity levels are

high. Wind speed was found to significantly reduce heat index values

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

by promoting evaporative cooling, while urban areas with limited

airflow experienced higher heat stress. Another key finding was the

impact of rainfall on heat index fluctuations—while rain initially

lowered temperatures, post-rain humidity spikes caused heat index

values to rebound quickly.

Additionally, atmospheric pressure changes influenced humidity

levels, affecting how heat was perceived over time. These insights

demonstrate the complexity of heat index prediction and the

importance of incorporating multiple meteorological variables beyond

just temperature and humidity.

Despite the model’s effectiveness, challenges were identified in

ensuring its consistency. The short data collection period (three

months) limited the ARIMA model’s ability to capture long-term

trends, making it more sensitive to sudden climate fluctuations.

Differences between data sources, particularly variations in sensor

calibrations from GlassStorm API and Wunderground, introduced

minor discrepancies in predictions. Additionally, while ARIMA

improved forecast accuracy compared to traditional methods, other

predictive techniques such as machine learning models

(LSTM, Random Forest) could further enhance reliability.

To improve the accuracy and consistency of heat index

forecasting, several recommendations can be made. Extending the

historical dataset to 6–12 months would allow ARIMA to better

identify seasonal trends and anomalies. Incorporating additional

variables such as solar radiation, cloud cover, and urban heat

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

island effects could refine predictions by accounting for localized

temperature variations. The integration of real-time weather API

updates would ensure dynamic adjustments to predictions based on

sudden changes in atmospheric conditions. Furthermore, using a

hybrid modeling approach that combines ARIMA with machine

learning techniques could enhance predictive capabilities by

capturing both linear and non-linear relationships in climate data.

Implementing automated data-cleaning methods would also help

reduce inconsistencies and improve model reliability.

Overall, the study demonstrates that data-driven heat index

prediction is essential for climate adaptation strategies,

particularly in regions prone to extreme heat events. By continuously

refining the forecasting model through improved data analysis

techniques, policymakers, weather agencies, and communities can

make more informed decisions to enhance public safety and

mitigate heat-related risks.

Chapter 3

Results and Discussion

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

Results

The analysis of the three datasets (Optimizing Heat Index

Prediction, GlassStorm (API), and Wunderground (Satellite))

provided critical insights into how different meteorological factors

influence the heat index. The results indicate:

1. Heat Index Variability Across Sources

o The heat index values differed slightly across the

three datasets, with variations in temperature,

humidity, and additional meteorological parameters

influencing the final predictions.

o GlassStorm (API) consistently reported lower heat

index values than the other sources, possibly due to

differences in sensor calibration or regional data

collection methods.

o Wunderground (Satellite) exhibited higher

fluctuations, indicating that satellite-based

measurements may be more sensitive to environmental

changes, particularly in localized conditions.

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

2. Temperature and Humidity Relationship

o The strongest correlation found was between

humidity and heat index, supporting the fact that high

humidity significantly amplifies perceived temperature.

o Even on days with moderate temperatures (78–82°F),

high humidity levels (above 85%) resulted in a heat

index exceeding 90°F, showing how relative humidity

affects human thermal perception.

o Days with lower humidity (below 75%) showed

smaller deviations in heat index, reinforcing the idea

that drier climates have less fluctuation in perceived heat

stress.

3. Wind Speed and Heat Index Reduction

o Wind speeds ranged from 2.6 mph to 5.6 mph across

the datasets, and their effects on the heat index were

substantial.

o Higher wind speeds were correlated with lower

heat index values, confirming that wind enhances

evaporative cooling, reducing thermal discomfort.

o Urban areas with low airflow showed higher heat

index values compared to rural locations with open

wind corridors. This highlights the role of city planning

and vegetation in mitigating heat stress.

4. Atmospheric Pressure’s Role

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

o A subtle yet consistent relationship was found between

lower atmospheric pressure and increased

humidity, leading to a higher heat index.

o Sudden drops in pressure often preceded spikes in

humidity, leading to heat index values that remained

elevated even as temperatures started declining in the

evening.

o Pressure fluctuations affected short-term heat

index predictions, suggesting that including pressure

trends in predictive models could enhance accuracy.

5. Precipitation and Post-Rainfall Effects

o Rain events initially led to a temporary cooling effect,

reducing the heat index as temperatures dropped and

cloud cover increased.

o However, post-rain humidity spikes caused a rapid

rebound in heat index values, particularly in low-wind

conditions.

o Heavy rainfall (>0.5 in) correlated with more

significant heat index reductions, while light rain

events (<0.2 in) had minimal impact.

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

Discussion

The findings confirm that heat index prediction models must

account for multiple meteorological factors beyond just temperature

and humidity. The discussion explores key interactions and their

implications:

1. The Limitations of Temperature-Based Heat Index Models

Traditional heat index models primarily rely on temperature and

humidity, but our analysis shows that this oversimplifies the

complexity of real-world heat perception.

 Inclusion of wind speed significantly refines predictions,

as it directly affects evaporative cooling. Current models that

exclude wind tend to overestimate heat index values in

breezy conditions and underestimate heat index during

stagnant air periods.

 Atmospheric pressure’s influence on humidity

fluctuations suggests that a static approach to heat

index modeling is insufficient. Including barometric pressure

trends can help predict spikes in humidity before they occur,

improving forecast reliability.

 The role of precipitation is often ignored in heat index

calculations, but our data show that post-rain humidity

surges can lead to misleading forecasts. Models that

assume rain always cools the environment may misrepresent

conditions in humid climates.

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

2. The Urban Heat Island (UHI) Effect on Heat Index

Our analysis detected noticeable differences in urban vs. rural

heat index readings, with urban areas consistently experiencing

higher values. Factors contributing to this include:

 Reduced wind circulation due to dense infrastructure,

limiting natural cooling effects.

 Increased heat retention from concrete and asphalt,

which store and re-emit heat even after sunset.

 Higher humidity levels from industrial and vehicular

emissions, compounding the effects of elevated temperatures.

 Limited vegetation and green spaces, reducing natural

cooling mechanisms such as transpiration.

Addressing UHI effects in heat index models requires

location-specific adjustments, incorporating data from smart

city weather stations to refine forecasts.

3. The Importance of Real-Time Weather Data for Heat Index

Prediction

One of the major challenges in heat index forecasting is the

dynamic nature of weather conditions. The datasets show that short-

term changes in wind speed, pressure, and precipitation

dramatically influence perceived heat but are often not

included in conventional prediction models.

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

 Real-time weather API integrations can improve forecast

accuracy by adjusting predictions based on live

meteorological inputs.

 AI-driven heat index predictors should use machine

learning models that continuously adapt to real-world

conditions, rather than relying solely on pre-defined

formulas.

 Regional climate differences must be factored into

models, as generic heat index calculations fail to

account for local weather patterns.

4. Future Considerations for Climate Adaptation and Public Health

As climate change intensifies, the frequency of extreme heat

events is expected to increase, making accurate heat index

forecasting more critical than ever.

 Improved heat index models will help emergency

response teams prepare for heatwaves, enabling targeted

interventions such as cooling stations in high-risk areas.

 Public awareness campaigns should educate

communities on the role of humidity, wind, and

precipitation in shaping real heat risk, preventing

misconceptions about temperature-based heat advisories.

 City planners can use heat index data to design more

heat-resilient infrastructure, such as ventilated urban

spaces, green roofs, and shaded public areas.

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

Final Thoughts

The study highlights significant opportunities to enhance heat

index prediction models by integrating wind speed, atmospheric

pressure, and precipitation data into forecasting algorithms. The

results demonstrate that real-world heat perception is shaped by

multiple meteorological factors, necessitating a multi-variable, AI-

driven approach to climate adaptation strategies. By improving

model accuracy, we can better prepare for rising global temperatures

and reduce the health risks associated with extreme heat events.

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

Chapter 4

SUMMARY OF FINDINGS, CONCLUSIONS AND RECOMMENDATIONS

Summary of Findings

The study analyzed heat index predictions from three sources:

Optimizing Heat Index Prediction, GlassStorm (API), and

Wunderground (Satellite). The key findings are:

1. Temperature and Humidity Impact

o Higher temperatures and humidity levels consistently

resulted in increased heat index values, confirming the

widely accepted correlation between these meteorological

factors.

o Across all datasets, humidity had a stronger

correlation with heat index fluctuations than

temperature alone. This indicates that even moderate

temperature increases can lead to a disproportionately

high heat index if humidity levels are also elevated.

o Variations in temperature and humidity across the

different data sources led to some deviations in predicted

heat index values, demonstrating the sensitivity of heat

index calculations to slight meteorological differences.

o Seasonal changes in humidity patterns also played

a significant role, with wetter seasons showing larger

fluctuations in the heat index compared to drier periods.

2. Influence of Wind Speed

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

o Wind speeds ranged from 2.6 mph to 5.6 mph across

the datasets, with notable effects on heat index values.

o Higher wind speeds were associated with lower heat

index values, as wind facilitates evaporative cooling and

reduces perceived heat stress. This effect is particularly

noticeable in conditions where humidity is high, as the

wind can counteract the retention of moisture on the skin.

o On the other hand, low wind speeds contributed to a more

stagnant environment, which exacerbated the effects of

high humidity, making conditions feel hotter than the

recorded temperature alone.

o Urban areas with limited airflow (e.g., due to

buildings or lack of vegetation) showed higher heat

index values compared to open rural areas.

3. Effect of Atmospheric Pressure

o Atmospheric pressure ranged from 29.6 in to 29.8 in,

with subtle but measurable effects on humidity and

temperature.

o While pressure variations were minor, a slight trend

suggested that lower pressure correlated with higher

humidity and increased heat index values. This

aligns with meteorological patterns where low-pressure

systems often bring increased moisture, leading to

warmer and more humid conditions.

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

o Higher atmospheric pressure was generally linked to drier

air, which slightly reduced the heat index despite high

temperatures.

o Sudden drops in pressure were associated with

incoming storm systems, which initially cooled the

environment but later led to increased humidity,

affecting the heat index.

4. Impact of Precipitation

o Rainfall events correlated with temporary reductions in

temperature and heat index values, particularly

during heavy precipitation periods where cloud cover and

evaporation contributed to localized cooling.

o However, after rainfall, increased humidity levels often

led to a rapid rebound in the heat index. This post-rain

humidity spike was most evident on days with minimal

wind, where moisture lingered in the air, intensifying

discomfort.

o The presence of precipitation data in forecasting models

could enhance accuracy by accounting for these sudden

fluctuations in heat index levels.

o Regions experiencing frequent short bursts of

rainfall tended to have fluctuating heat index

values compared to areas with longer, sustained

rainfall.

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

Conclusion

The accuracy of heat index prediction is influenced by multiple

meteorological variables beyond just temperature and humidity.

Specifically:

 Wind Speed: Higher wind speeds help dissipate heat and

reduce the perceived temperature, leading to a lower heat

index. This effect is more pronounced in humid environments

where wind accelerates evaporative cooling. However, in

areas with obstructed airflow, such as dense urban

centers, the cooling effect of wind is significantly

reduced.

 Rainfall: While rainfall initially cools the environment by

lowering ambient temperatures, post-rain humidity spikes

can cause the heat index to rise quickly, leading to

deceptive cooling effects that may not persist for long periods.

The timing and duration of rainfall events need to be

accounted for to refine prediction accuracy.

 Atmospheric Pressure: Lower pressure systems are often

associated with increased humidity, which in turn elevates the

heat index. This means that areas experiencing fluctuating

pressure levels may see inconsistent heat index predictions if

this factor is not incorporated into the model.

Given these interactions, models relying only on temperature and

humidity without factoring in wind speed, precipitation, and

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

pressure tend to overestimate or underestimate the actual heat

index values, leading to potential inaccuracies in climate adaptation

strategies.

Recommendations

To optimize and improve heat index prediction models, the

following steps are recommended:

1. Incorporate Additional Meteorological Variables

o Enhance the model by integrating wind speed,

atmospheric pressure, and precipitation data for

better accuracy. This will allow predictions to be adjusted

based on real-time atmospheric changes.

o Utilize real-time weather API data to dynamically update

heat index values, ensuring that sudden shifts in weather

conditions are reflected in forecasts.

o Include urban heat island effects as an additional

factor in modeling heat index variations in city

environments.

2. Machine Learning and Hybrid Modeling

o Implement machine learning models, such as Random

Forest or Neural Networks, to capture complex

patterns and improve predictive accuracy.

o Hybrid models combining regression analysis with AI-

driven prediction algorithms can improve forecast

reliability by leveraging the strengths of both traditional

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

statistical models and adaptive machine learning

techniques.

o Consider using deep learning approaches such as Long

Short-Term Memory (LSTM) networks to capture

temporal dependencies in climate patterns.

o Use ensemble learning techniques that combine

multiple models to increase prediction robustness.

3. Data Preprocessing and Outlier Removal

o Implement automated outlier detection algorithms to

identify and correct anomalies in meteorological data,

preventing extreme values from distorting predictions.

o Apply cross-validation techniques to improve

prediction stability and generalizability, ensuring that

models do not overfit to specific historical datasets.

o Use interpolation and imputation techniques to handle

missing weather data, reducing the impact of incomplete

datasets on forecasting accuracy.

o Apply feature engineering techniques to better

capture interactions between meteorological

variables, improving predictive insights.

4. Fine-Tune Regression Models

o Use polynomial regression instead of simple linear

models to better capture nonlinear relationships between

temperature, humidity, and heat index.

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

o Conduct periodic recalibrations using updated datasets

to adapt to seasonal variations and evolving climate

conditions.

o Evaluate the effectiveness of different regression

techniques, such as ridge regression and Bayesian

regression, to improve model robustness.

o Test different hyperparameters within regression

models to optimize performance and reduce

forecasting errors.

By integrating these enhancements, the heat index predictor

can provide more reliable and real-time forecasts, helping

communities, weather agencies, and policymakers implement more

effective climate adaptation strategies. These improvements will

ensure that public safety measures, such as early heat advisories and

emergency response protocols, are based on the most accurate and

comprehensive predictions available.

REFERENCES

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

Dahl, K., Licker, R., Abatzoglou, J. T., & Declet-Barreto, J. (2019). Increased

frequency of and population exposure to extreme heat index days in the United

States during the 21st century. Environmental Research Communications, 1(7),

075002. https://doi.org/10.1088/2515-7620/ab27cf

Hewage, P., Trovati, M., Pereira, E., & Behera, A. (2020). Deep learning-based

effective fine-grained weather forecasting model. Pattern Analysis and

Applications, 24(1), 343–366. https://doi.org/10.1007/s10044-020-00898-1

Khatana, S. A. M., Werner, R. M., & Groeneveld, P. W. (2022). Association of

extreme heat and cardiovascular mortality in the United States: A county-level

longitudinal analysis from 2008 to 2017. Circulation, 146(3), 249–261.

https://doi.org/10.1161/circulationaha.122.060746

Romps, D. M. (2024). Heat index extremes increasing several times faster than

the air temperature. Environmental Research Letters, 19(4), 041002.

https://doi.org/10.1088/1748-9326/ad3144

Wang, S., Wu, C. Y. H., Richardson, M. B., Zaitchik, B. F., & Gohlke, J. M. (2021).

Characterization of heat index experienced by individuals residing in urban and

rural settings. Journal of Exposure Science & Environmental Epidemiology, 31(4),

641–653. https://doi.org/10.1038/s41370-021-00303-x

Chen, Y.-C. (2023). Thermal indices for climate and perception evaluation. Journal

of Building Engineering, 74, 106886.

Miranda, F. M., Oliveira, A. L. I., & Freitas, A. A. (2023). A Python Package for

Hierarchical Machine Learning. SoftwareX, 21, 101344.

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

Polasky, A. D., Lora, C., & Vera, C. S. (2023). CCdownscaling: A Python package

for statistical downscaling of climate data. Environmental Modelling & Software,

167, 105779.

Nailman, A. (2024, May 28). Comparing ML and statistical models: effectiveness

and performance. Machine Learning Models.

https://machinelearningmodels.org/comparing-ml-and-statistical-models-

effectiveness-and-performance/

Singh, H. (2024, December 20). 6 Assumptions of Linear Regression. Analytics

Vidhya. Retrieved

from https://www.analyticsvidhya.com/blog/2016/07/deeper-regression-analysis-

assumptionsplots-solutions/

Addagatla, N. (2023, October 6). Which Machine Learning Algorithm to Use

When? with RealLife Examples. Medium. Retrieved from

https://medium.com/%40nitesh.py/which-machinelearning-algorithm-to-use-

when-with-real-life-examples-7090c65fdab1

Gupta, S. (2021, December 13). Pros and cons of various Machine Learning

algorithms. Medium. https://medium.com/towards-data-science/pros-and-cons-of-

various-classification-ml-algorithms-3b5bfb3c87d6

Breton, D. (2022, January 6). A full guide on choosing the right machine learning

algorithm. Medium. https://medium.com/@davidbreton03/a-full-guide-on-

choosing-the-right-machine-learning-algorithm-5fa282a0b2a1

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

Sourjah, F.S., & Pemarathne, W.P.J. (2022). Heat Wave Prediction using Machine

Learning Techniques: A Review. International Journal of Computer Applications,

184(16). Retrieved from

https://www.ijcaonline.org/archives/volume184/number16/32404-2022922162/

Dutchen, S. (2024, Autumn). Machine Learning Can Predict the Weather — and

Human Health. Harvard Medical School. Retrieved from

https://magazine.hms.harvard.edu/articles/machinelearning-can-predict-weather-

and-human-health

Boudreault, J., Ruf, A., Campagna, C., & Chebana, F. (2024). Multi-region models

built with machine and deep learning for predicting. Sustainable Cities and

Society, 105785. Retrieved from https://ouci.dntb.gov.ua/en/works/4OgGQap4/

Arnold, J., & Arnold, J. (2024, June 14). Enhancing Emergency Detection and Early

Warning Response: The Role of AI and Machine Learning in Severe Weather

Preparedness | HQE Systems. HQE Systems | Veteran Owned.

https://hqesystems.com/enhancing-emergency-detection-and-early-warning-

response-the-role-of-ai-and-machine-learning-in-severe-weather-preparedness/

Maitra, M. (2023, July 26). Assumptions in regression: why, what, and how.

DATAVERSITY. https://www.dataversity.net/assumptions-in-regression-why-what-

and-how/Ijraset. (2021.). Climate Change Forecasting using Machine Learning

Algorithms. IJRASET. https://www.ijraset.com/research-paper/climate-change-

forecasting-using-ml-algorithms

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

An Introduction to the Statistical Modelling of Climate Change: 1850-2020.

(2021). University of Birmingham. Retrieved from

https://www.birmingham.ac.uk/research/climate/climatepublications/nature/an-

introduction-to-the-statistical-modelling-of-climate-change-1850-2020

Renoult, M., Annan, J. D., Hargreaves, J. C., Sagoo, N., Flynn, C., Kapsch, M., Li, Q.,

Lohmann, G., Mikolajewicz, U., Ohgaito, R., Shi, X., Zhang, Q., & Mauritsen, T.

(2020). A Bayesian framework for emergent constraints: case studies of climate

sensitivity with PMIP. Climate of the Past, 16(5), 1715–1735.

https://doi.org/10.5194/cp-16-1715-2020

Levi, D., Gispan, L., Giladi, N., & Fetaya, E. (2019, May 28). Evaluating and

calibrating uncertainty prediction in regression tasks. arXiv.org.

https://arxiv.org/abs/1905.11659

Trenary, L., & DelSole, T. (2022). Advancing interpretability of machine-learning

prediction models. Environmental Data Science, 1, e14.

Lipton, Z. C. (2016, June 10). The Mythos of Model Interpretability. arXiv.org.

https://arxiv.org/abs/1606.03490

Niculae, A., Oprea, S., Văduva, A., Bâra, A., & Andreescu, A. (2024). Assessing the

role of machine learning in climate research publications. Sustainability, 16(24),

11086. https://doi.org/10.3390/su162411086

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

De Burgh-Day, C. O., & Leeuwenburg, T. (2023). Machine learning for numerical

weather and climate modelling: a review. Geoscientific Model Development,

16(22), 6433–6477. https://doi.org/10.5194/gmd-16-6433-2023

Niculae, A., Oprea, S., Văduva, A., Bâra, A., & Andreescu, A. (2024b). Assessing

the Role of Machine Learning in Climate Research Publications. Sustainability,

16(24), 11086. https://doi.org/10.3390/su162411086

Climate Technology Progress Reports - UNEP-CCC. (2025, January 20). UNEP-CCC.

https://unepccc.org/climate-technology-progress-reports/

Ten New Insights in Climate Science 2023. (2023). Cambridge Global

Sustainability Journal. Retrieved from

https://www.cambridge.org/core/journals/global-sustainability/article/ten-

newinsights-in-climate-science-2023/F7F1C10C07FD241BFE30ACC4BA555A56

Senevirathne, J. (2024, November 4). Building a Weather Prediction Model with

Machine Learning: A Step-by-Step Guide. Medium.

https://janaksenevirathne.medium.com/building-a-weather-prediction-model-

with-machine-learning-a-step-by-step-guide-9eaf768171be

De Burgh-Day, C. O., & Leeuwenburg, T. (2023b). Machine learning for numerical

weather and climate modelling: a review. Geoscientific Model Development,

16(22), 6433–6477. https://doi.org/10.5194/gmd-16-6433-2023

Climate change: improved prediction of heatwaves thanks to AI | CNRS. (2023,

April 3). CNRS. https://www.cnrs.fr/en/press/climate-change-improved-prediction-

heatwaves-thanks-ai

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

Energy end-use data collection methodologies and the emerging role of digital

technologies – Analysis - IEA. (2020, October 1). IEA.

https://www.iea.org/reports/energy-end-use-data-collection-methodologies-and-

the-emerging-role-of-digital-technologies

Interpretable Machine Learning for Weather and Climate. (2022). MDPI

Sustainability. Retrieved from https://www.mdpi.com/2071-1050/16/24/11086

Niculae, A., Oprea, S., Văduva, A., Bâra, A., & Andreescu, A. (2024c). Assessing

the Role of Machine Learning in Climate Research Publications. Sustainability,

16(24), 11086. https://doi.org/10.3390/su162411086

How can Artificial Intelligence Help Improve Climate Forecasting and Risk

Information? (n.d.). ICPAC. https://www.icpac.net/news/how-can-artificial-

intelligence-help-improve-climate-forecasting-and-risk-information/

Know Your Climate Bias. (2020.). Center for Leadership in Global Sustainability |

Virginia Tech. https://cligs.vt.edu/blog/know-your-climate-bias.html

Learning to Correct Climate Projection Biases | https://eesm.science.energy.gov/.

(n.d.). https://eesm.science.energy.gov/research-highlights/learning-correct-

climate-projection-biases

Usgcrp. (n.d.). Climate Science Special Report.

https://science2017.globalchange.gov/chapter/appendix-a/

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

Barton, G., Lee, N. T., & Resnick, P. (2019, May 22). Algorithmic bias detection

and mitigation: Best practices and policies to reduce consumer harms. Brookings.

https://www.brookings.edu/articles/algorithmic-bias-detection-and-mitigation-

best-practices-and-policies-to-reduce-consumer-harms/

GitHub - Missing Datasets. (2021). GitHub Repository. Retrieved from

https://github.com/MimiOnuoha/missing-datasets

Closing Climate and Disaster Data Gaps: New Challenges, New Thinking. (2022).

United Nations Office for Disaster Risk Reduction (UNDRR). Retrieved from

https://www.undrr.org/publication/closing-climate-and-disaster-data-gaps-new-

challenges-newthinking

METEO 810: Weather and Climate Data Sets. (2023). Pennsylvania State

University. Retrieved from

https://www.e-education.psu.edu/meteo810/content/l5_p4.html

Sensor Calibration Techniques for Improved Reliability and Precision. (2023).

Sensor Networks Journal. Retrieved from https://sensor-networks.org/sensor-

calibration-techniques-for-improvedreliability-and-precision/

Essential Sensor Calibration Techniques for Accurate Measurements. (2021). The

Tech Artist. Retrieved from https://thetechartist.com/sensor-calibration-

techniques/ Specifying Calibration of Environmental Sensors. (2023). Pacific

Northwest National Laboratory (PNNL). Retrieved from

https://www.pnnl.gov/publications/specifying-calibrationenvironmental-sensors

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

Sensor Calibration and Characterization to Meet Climate Monitoring

Requirements. (2022). National Institute of Standards and Technology (NIST).

Retrieved from

https://www.nist.gov/publications/sensor-calibration-and-characterization-meet-

climatemonitoring-requirements

Improving Climate Data Records with FIDUCEO. (2023). European Organisation

for the Exploitation of Meteorological Satellites (EUMETSAT). Retrieved from

https://www.eumetsat.int/features/improving-climate-data-records-fiduceo

Fidelity and Uncertainty in Climate Data Records. (2024). Hamburg University

Meteorological Institute. Retrieved from

https://www.mi.uni-hamburg.de/en/arbeitsgruppen/strahlung

undfernerkundung/projektarchiv/fiduceo.html

Heat Hazards and Migrant Rights. (2023). Federation of American Scientists

(FAS). Retrieved from https://fas.org/publication/heat-hazards-and-migrant-rights/

Industry Partnerships Create Policy Tools to Protect Outdoor Laborers in Extreme

Heat. (2023).University of Connecticut. Retrieved from

https://today.uconn.edu/2023/08/industry-partnershipscreate-policy-tools-to-

protect-outdoor-laborers-in-extreme-heat/

Turner Studying Effects of Rising Temperatures on Construction Workers. (2023).

Turner Construction Company. Retrieved from

https://www.turnerconstruction.com/insights/turnerstudying-effects-of-rising-

temperatures-on-construction-workers Rising Heat Risks for Workers. (2023).

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

Climate Central. Retrieved from

https://www.climatecentral.org/climate-matters/rising-heat-risks-for-workers

Feeling the Heat – The Effects of Higher Temperatures on People, Crops, and

Livelihoods. (2023). CGIAR Research Program on Climate Change, Agriculture and

Food Security. Retrieved from

https://adaptationatlas.cgiar.org/quick-reads/feeling-the-heat

The Impact of High Temperatures on Livestock. (2023). Animal Agriculture and

Climate Change. Retrieved from

https://www.animalagricultureclimatechange.org/high-temperatures-livestock/

Quantifying the Impact of Future Extreme Heat on the Outdoor Work Sector in the

United States. (2023). Global Heat Health Information Network (GHHIN).

Retrieved from https://ghhin.org/resources/quantifying-the-impact-of-future-

extreme-heat-on-the-outdoor-worksector-in-the-united-states/

2022 Young Farmers Climate and Water Policy Recommendations. (2022).

National Young Farmers Coalition. Retrieved from

https://www.youngfarmers.org/2022/08/2022-young-farmersclimate-and-water-

policy-recommendations/

11 Ways Farmers Are Adapting to the Unpredictability of Climate Change. (2023).

Earth.Org. Retrieved from https://earth.org/11-ways-farmers-are-adapting-to-the-

unpredictability-of-climatechange/

The School Year Is Getting Hotter: How Does Heat Affect Student Learning and

Well-being?.

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

(2022). Education Week. Retrieved from https://www.edweek.org/leadership/the-

school-year-isgetting-hotter-how-does-heat-affect-student-learning-and-well-

being/2022/09

Heatwaves Pose Health Risks, Disrupt Students’ Learning. (2023). Save the

Children Philippines. Retrieved from https://journal.com.ph/heatwaves-pose-

health-risks-disrupt-students-learningsays-save-the-children-philippines/

Outdoor Workers Face Increased Vulnerability from Weather. (2023). Labor Press.

Retrieved from https://www.laborpress.org/outdoor-workers-face-increased-

vulnerability-from-weather/

Protecting Agricultural Workers in a Changing Climate. (2023). Union of

Concerned Scientists. Retrieved from

https://www.ucsusa.org/resources/protecting-agricultural-workers-

changingclimate

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

APPENDIX A

LETTERS OF ACKNOWLEDGEMENT

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

APPENDIX B

DOCUMENTATIONS

Our adviser chatted with Gio and agreed that we will proceed

with the MIP category for our research.

The other members came to our house to encode and gather

data, allowing us to slowly work on our research.

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

This was when we went to PAGASA to ask for data, and they also taught us

how to obtain heat index information through the website they

recommended.

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

Kuya Zoed and his friend visited our school to help us with coding and to

guide us through other aspects of our research.

Our other members visited PAGASA Surigao Station to deliver the

acknowledgment paper and have it signed.

PAGE \* MERGEFORMAT i
Surigao del Norte National High School
Peñaranda St., Surigao City

APPENDIX C
CODING

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

CURRICULUM VITAE
I. Personal Details

Name: Mae Ann J. Cabug-os

Contact Number: 09813519540

Email Address: [email protected]

Age: 17

Gender: Female

Date of Birth: June 24, 2007

Place of Birth: Western Bicutan

Home Address: BRGY. Togbongon P-1 Surigao City

Citizenship: Filipino

Religion: Roman Catholic Church

Civil Status: Single

Name of Father: Antony P. Cabug-os

Name of Mother: Ma. Angelita J. Cabug-os

II. Educational Attainment

Secondary: Timamana National High School

2018-2022

Elementary: San Isidro Elementary School

2011-2018

PAGE \* MERGEFORMAT i
Surigao del Norte National High School
Peñaranda St., Surigao City

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

CURRICULUM VITAE

Personal Details

Name: Alexis D. Gio

Contact Number: 09461004515

Email Address: [email protected]

Age: 18

Gender: Male

Date of Birth: November 08, 2006

Place of Birth: Caraga Regional Hospital Surigao City

Home Address: Borromeo St. Surigao City

Citizenship: Filipino

Religion: Roman Catholic Church

Civil Status: Single

Name of Father: Roger C. Gio

Name of Mother: Elizabeth D. Gio

II. Educational Attainment

Secondary: Surigao City National High School

2018-2022

Elementary: C.V Diez Memorial Central Elementary School

2011-2018

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

CURRICULUM VITAE

Personal Details

Name: Irene Jade D. Nono

Contact Number: 09308353222

Email Address: [email protected]

Age: 17

Gender: Female

Date of Birth: June 11, 2007

Place of Birth: Mainit, Surigao City

Home Address: BRGY. Bonifacio, Acean Heights P-1 Surigao City

Citizenship: Filipino

Religion: Christian

Civil Status: Single

Name of Father: Erryl James N. Nono

Name of Mother: Lineth D. Nono

II. Educational Attainment

Secondary: Jesus Cares Foundational Learning Academy

2018-2022

Elementary: Bonifacio Elementary School

2011-2018

PAGE \* MERGEFORMAT i
 Surigao del Norte National High School
Peñaranda St., Surigao City

CURRICULUM VITAE

Personal Details

Name: Charise Villazon Orjela

Contact Number: 09466752969

Email Address: [email protected]

Age: 18

Gender: Female

Date of Birth: February 11, 2007

Place of Birth: Caraga Regional Hospital Surigao City

Home Address: Surigao City

Citizenship: Filipino

Religion: Roman Catholic Church

Civil Status: Single

Name of Father: Rimmy Siscon Orjela

Name of Mother: Rebecca Villazon Orjela

II. Educational Attainment

Secondary: Lakandula National High School

2018-2022

Elementary: Lakandula Elementary School

2011-2018

PAGE \* MERGEFORMAT i
Surigao del Norte National High School
Peñaranda St., Surigao City

PAGE \* MERGEFORMAT i