Periodic Function Modeling — AHL 3.8 Further trigonometric modelling (periodic functions in context) (HL extension) - IB Mathematics Applications and Interpretation

SECTION 1

Historical Context & Motivation

Humans have tracked repeating patterns in nature for millennia — from the predictable swing of the seasons to the ebb and flow of ocean tides. Ancient civilizations in Babylon and Egypt recorded these cycles to plan agriculture, navigation, and religious festivals. Yet it was not until mathematicians developed trigonometric functions that we gained a precise language for describing any phenomenon that repeats at regular intervals. The journey from star-gazing priests to modern climate scientists modeling temperature data is fundamentally a story of periodic function modeling.

~150 CE • Ptolemy's Chord Tables

Claudius Ptolemy compiled tables of chords — the ancient precursor to sine values — to model planetary motion, one of the earliest attempts to describe periodic celestial cycles mathematically.

1822 • Fourier's Theorem

Joseph Fourier demonstrated that any periodic function can be expressed as a sum of sine and cosine waves. This breakthrough showed that sinusoidal models are universal building blocks for all repeating patterns.

1898 • Tidal Prediction Machines

Lord Kelvin's tide-predicting machines used mechanical gears to sum multiple sinusoidal components, producing accurate tide charts for harbours worldwide — applied periodic modeling in action.

1965 • Fast Fourier Transform (FFT)

Cooley and Tukey published the FFT algorithm, enabling computers to decompose complex signals into sinusoidal components rapidly. This revolutionised everything from audio processing to medical imaging.

Today • Data-Driven Periodic Models

Scientists and engineers routinely fit sinusoidal regression models to real-world data — temperature records, electricity demand, biological rhythms — making periodic function modeling an essential skill in the IB Applications and Interpretation course.

The central question driving this topic is deceptively simple: given a set of data that rises and falls in a repeating pattern, how do we write a single function that captures its behaviour and allows us to make predictions? In AHL 3.8 you will learn to answer this by constructing sinusoidal models whose parameters — amplitude, period, phase shift, and vertical shift — are determined from context rather than handed to you in a textbook exercise.

SECTION 2

Core Principles & Definitions

Before diving into contextual problems, you need a solid grasp of the four parameters that shape every sinusoidal curve. Each one has a clear real-world meaning, and understanding them lets you move fluently between a graph, an equation, and the physical situation it describes.

Amplitude (a)

The amplitude measures how far the function swings above or below its central axis. It equals half the distance between the maximum and minimum values: a = (max − min) / 2.

Period (T) and b-value

The period is the length of one complete cycle. It relates to the coefficient b by T = 2π / b. A shorter period means more cycles packed into the same horizontal span.

Phase Shift (c / b)

The phase shift slides the curve left or right along the horizontal axis. It tells you where in the cycle the function starts. In context, it aligns the model with the timing of real events.

Vertical Shift (d)

The vertical shift (also called the midline or principal axis) raises or lowers the entire curve. It equals the average of the maximum and minimum values: d = (max + min) / 2.

✦

KEY TAKEAWAY

Think of a sinusoidal function like a wave on a jump rope. The amplitude is how high you swing the rope; the period is how quickly you shake it; the phase shift is whether you start at the top, bottom, or middle of a shake; and the vertical shift is the height at which you hold your hand. Change any one, and the wave looks different — but it is still the same fundamental shape.

SECTION 3

Visual Explanation — Anatomy of a Sinusoidal Model

The diagram shows one full cycle of y = a sin(b(x − c)) + d. The cyan curve traces the sinusoidal wave. The pink dashed line marks the amplitude from midline to maximum, while the violet bracket spans one full period T along the horizontal axis. The dashed horizontal lines represent the maximum, midline (d), and minimum.

Notice how the entire curve is symmetric about the horizontal midline at y = d. The maximum value is d + a, and the minimum value is d − a. In a real-world context, you will often read the maximum and minimum directly from data (for example, the hottest and coldest monthly temperatures), and then compute a and d from them. The period T is determined by how long one full cycle takes — a full year for seasonal data, roughly 12.4 hours for ocean tides, or 24 hours for daily temperature swings.

SECTION 4

Mathematical Framework

The general sinusoidal model used throughout AHL 3.8 takes two equivalent forms. Both are given in the IB formula booklet, so you should be comfortable working with either one.

GENERAL SINE MODEL

y = a sin(b(x − c)) + d

a = amplitude | b = 2π / T | c = phase shift (horizontal translation) | d = vertical shift (midline)

GENERAL COSINE MODEL

y = a cos(b(x − c)) + d

Cosine is identical to sine but shifted by π/(2b) horizontally. Choose whichever aligns more naturally with your data — cosine starts at a maximum, sine starts at the midline going up.

PARAMETER FORMULAS FROM DATA

a = (max − min) / 2 d = (max + min) / 2 b = 2π / T

max and min are the largest and smallest observed y-values. T is the time for one complete cycle. The phase shift c is found by considering when the function reaches its maximum (for cosine) or crosses the midline going upward (for sine).

A critical detail for IB exams: when the problem specifies angles in degrees, the b-value formula changes to b = 360° / T instead of 2π / T. Always check whether the context uses time (hours, months) with radian measure or degree measure — the IB occasionally mixes both. Your GDC's angle mode must match.

DEGREE-MODE B-VALUE

b = 360° / T

Use this when the horizontal axis is measured in units like months or hours and you want the argument in degrees.

💡

IB Exam Tip

When the question says 'Find a sinusoidal model', you must state the equation with numerical values for a, b, c, and d, and define what x and y represent in context (e.g. x = month number, y = average temperature in °C). Full marks require interpretation of parameters.

SECTION 5

Extracting Parameters from Real-World Data

The distinguishing feature of AHL 3.8 is that you must extract model parameters from a given context — a table of data, a description of a physical system, or a graph. The process follows a consistent sequence, and practising it with varied contexts is the best preparation for the exam.

This flowchart summarises the six-step process for constructing a sinusoidal model from contextual data. The bottom panel applies the process to monthly temperature data, showing how each parameter is determined from the given values.

A common point of confusion is choosing between sine and cosine. The practical rule is straightforward: if you know when the maximum occurs, cosine is more convenient because cos(0) = 1, so the maximum happens when the argument equals zero, giving c = x-value of the max directly. If you prefer sine, note that a maximum of the standard sine function occurs at π/2, so you need to account for that quarter-period offset. Either choice is acceptable on the IB exam; just be consistent and show your reasoning.

⚠

Negative a-values

If the data starts at a minimum rather than a maximum, you can use a negative amplitude (a < 0) with cosine, or adjust the phase shift. For example, if the minimum is at x = 1, then y = −a cos(b(x − 1)) + d places a trough at x = 1. Both approaches are valid.

SECTION 6

Worked Example — Tidal Heights

A harbour records the following tidal data: high tide occurs at 03:00 with a water depth of 4.8 m, and low tide occurs at 09:15 with a depth of 1.2 m. The tides follow a sinusoidal pattern. Find a model of the form y = a cos(b(t − c)) + d, where t is the time in hours after midnight and y is the depth in metres. Then use the model to predict the water depth at 17:00.

Modelling Tidal Heights

1
Step 1 — Identify Maximum and Minimum

From the problem: maximum depth = 4.8 m at t = 3 (03:00), minimum depth = 1.2 m at t = 9.25 (09:15, since 15 min = 0.25 h).

max = 4.8 m, min = 1.2 m
2
Step 2 — Calculate Amplitude and Vertical Shift

a = (max − min) / 2 = (4.8 − 1.2) / 2 = 3.6 / 2 = 1.8. The vertical shift is d = (max + min) / 2 = (4.8 + 1.2) / 2 = 6.0 / 2 = 3.0.

a = 1.8 m, d = 3.0 m
3
Step 3 — Determine the Period

The time from high tide to the next low tide is half a period. From t = 3 to t = 9.25 is 6.25 hours, so the half-period is 6.25 h, giving a full period of T = 2 × 6.25 = 12.5 h. This is consistent with the approximately 12.4-hour tidal cycle.

T = 12.5 hours
4
Step 4 — Compute the b-value

b = 2π / T = 2π / 12.5 = 4π / 25 ≈ 0.5027.

b = 4π/25 ≈ 0.503
5
Step 5 — Find the Phase Shift

Since we are using cosine and the maximum occurs at t = 3, the phase shift c = 3. This means cos(b(t − 3)) reaches its maximum value of 1 when t = 3, as required.

c = 3
6
Step 6 — Write the Model and Predict

The model is y = 1.8 cos(4π/25 × (t − 3)) + 3.0. To find the depth at 17:00 (t = 17): y = 1.8 cos(4π/25 × (17 − 3)) + 3 = 1.8 cos(4π/25 × 14) + 3 = 1.8 cos(56π/25) + 3. Now 56π/25 = 2.24π. Since cos(2.24π) = cos(0.24π) ≈ cos(43.2°) ≈ 0.7265, we get y ≈ 1.8 × 0.7265 + 3 ≈ 1.308 + 3 = 4.31 m.

y = 1.8 cos(4π/25 (t − 3)) + 3.0; depth at 17:00 ≈ 4.31 m

💡

Check Your Answer

The predicted depth of 4.31 m at 17:00 makes physical sense. Since one full period is 12.5 hours after the first high tide (03:00), the next high tide should be at approximately 15:30, and 17:00 is only 1.5 hours past that peak. The water is still close to its maximum — consistent with 4.31 m being close to the maximum of 4.8 m.

SECTION 7

Strengths and Limitations of Sinusoidal Models

Sinusoidal models are incredibly versatile, but they are not perfect representations of every periodic phenomenon. Understanding when a sinusoidal model is appropriate — and when it might fall short — is essential for the IB exam, where you may be asked to discuss the model's validity.

Aspect	Strengths	Limitations
Simplicity	Only four parameters (a, b, c, d) to determine. Easy to fit from a small dataset or even just the max, min, and timing.	Assumes perfect symmetry between the rising and falling parts of each cycle. Many real-world phenomena (e.g. heartbeats, sawtooth waves) are asymmetric.
Prediction	Extrapolates naturally: plug in any future (or past) x-value and get a prediction. Excellent for short-term forecasting of tides, temperatures, daylight hours.	Assumes the pattern repeats identically forever. Does not account for trends (e.g. global warming adding an upward drift to temperature data).
Mathematical tractability	Derivatives and integrals of sine/cosine are well-known. Easy to find rates of change (useful for HL calculus connections).	A single sine/cosine may not fit data with multiple overlapping periods (e.g. tides influenced by both the moon and sun). Fourier sums may be needed.
Fit quality	GDC sinusoidal regression (sinReg) can optimise parameters for the best least-squares fit to scattered data.	Regression may converge on unexpected parameter values if the data is noisy or if fewer than two full cycles are present.

Strengths vs. Limitations of Sinusoidal Models

✦

KEY TAKEAWAY

A sinusoidal model is like using a perfectly round wheel to represent a shape that is almost round. It captures the main pattern — the regular rise and fall — beautifully. But if the real shape has bumps, flat spots, or a gradual drift, you may need to combine the sinusoidal model with other functions (like a linear trend) to get a closer fit.

SECTION 8

Connections to Advanced Theory

AHL 3.8 periodic models connect naturally to several other parts of the IB syllabus and to mathematics beyond the course. Recognising these links deepens your understanding and prepares you for cross-topic exam questions.

AHL 3.8 Concept	Advanced Connection
Finding rate of change at a point	AHL 5.9 — Differentiation of trig functions. The derivative of a sin(b(x − c)) + d is ab cos(b(x − c)), giving the instantaneous rate of change (e.g. how fast the temperature is rising in April).
Area under one cycle	AHL 5.11 — Integration of trig functions. Integrating the model over one period gives accumulated quantities (e.g. total heating degree-hours).
Sinusoidal regression on GDC	AHL 4.12 — Non-linear regression. Using sinReg on your GDC fits a, b, c, d simultaneously by minimising the sum of squared residuals — an optimisation technique.
Adding a linear trend to a periodic model	Composite functions: y = mx + a sin(b(x − c)) + d models phenomena like rising average temperatures with a superimposed seasonal cycle. This is a real technique used in climate science.

How AHL 3.8 connects to the wider IB syllabus and beyond

Looking further ahead, Fourier analysis — studied in university mathematics and engineering — generalises the single-sinusoid model to sums of many sinusoids, each with its own frequency and amplitude. This is how audio engineers decompose complex sounds into individual notes, and how MRI machines reconstruct images of the human body. The single sinusoidal model you learn in AHL 3.8 is the essential first step on that road.

SECTION 9

Practice Problems

Question 1: CONCEPTUAL

A student claims that the function y = 5 sin(2x) + 10 has a range of [5, 15]. Another student says the range is [−5, 5]. Which student is correct, and why? Explain in terms of the parameters a and d.

Show answer

The first student is correct. The amplitude a = 5 and the vertical shift d = 10. The sine function oscillates between −1 and 1, so y ranges from 5(−1) + 10 = 5 to 5(1) + 10 = 15. The range is [5, 15]. The second student forgot to account for the vertical shift d = 10, which raises the midline from y = 0 to y = 10.

Question 2: BASIC CALCULATION

The depth of water in a harbour is modelled by h(t) = 3 cos(π/6 × t) + 5, where h is in metres and t is in hours after midnight. Find: (a) the amplitude, (b) the period, (c) the depth at t = 4 hours.

Show answer

(a) The amplitude is a = 3 m, read directly from the coefficient of cosine. (b) The period is T = 2π / b = 2π / (π/6) = 12 hours. (c) At t = 4: h(4) = 3 cos(π/6 × 4) + 5 = 3 cos(2π/3) + 5 = 3 × (−0.5) + 5 = −1.5 + 5 = 3.5 m.

Question 3: INTERMEDIATE

The average monthly temperature in a city reaches a maximum of 28 °C in August (month 8) and a minimum of 6 °C in February (month 2). Assuming a sinusoidal model of the form T(x) = a cos(b(x − c)) + d, where x is the month number (January = 1), find a, b, c, and d, and write the full equation.

Show answer

Amplitude: a = (28 − 6)/2 = 11. Vertical shift: d = (28 + 6)/2 = 17. Period: T = 12 months (one year), so b = 2π/12 = π/6. Phase shift: since the maximum occurs at x = 8, we set c = 8. The model is T(x) = 11 cos(π/6 (x − 8)) + 17. To verify: T(8) = 11 cos(0) + 17 = 28 ✓. T(2) = 11 cos(π/6 × (−6)) + 17 = 11 cos(−π) + 17 = 11(−1) + 17 = 6 ✓.

Question 4: APPLIED

A Ferris wheel has a diameter of 40 m and its centre is 25 m above the ground. It completes one full revolution every 8 minutes. A rider boards at the lowest point at t = 0. (a) Write a sinusoidal model h(t) for the rider's height above the ground. (b) Find the first time the rider is 40 m above the ground.

Show answer

(a) The radius is 20 m, so the maximum height = 25 + 20 = 45 m and the minimum height = 25 − 20 = 5 m. Thus a = 20 and d = 25. The period T = 8 min, so b = 2π/8 = π/4. Since the rider starts at the lowest point (minimum) at t = 0, we use h(t) = −20 cos(π/4 × t) + 25. Check: h(0) = −20(1) + 25 = 5 m ✓ (bottom). (b) Set 40 = −20 cos(πt/4) + 25, so −20 cos(πt/4) = 15, giving cos(πt/4) = −0.75. Then πt/4 = cos⁻¹(−0.75) ≈ 2.4189, so t ≈ 4 × 2.4189 / π ≈ 3.08 minutes.

Question 5: CRITICAL THINKING

A researcher models yearly average temperature in a city as T(x) = 11 cos(π/6 (x − 7)) + 15, where x is the month number. However, due to climate change, the average temperature is increasing by 0.03 °C per year. Propose a modified model T(x, n) that accounts for both the seasonal cycle and the long-term warming trend, where n is the number of years since the base year. What assumptions does your model make, and what are its limitations?

Show answer

A suitable model is T(x, n) = 11 cos(π/6 (x − 7)) + 15 + 0.03n. This adds a linear trend 0.03n to the vertical shift, raising the midline by 0.03 °C each year. The model assumes: (1) the warming rate is constant over time (linear, not exponential), (2) the amplitude and phase of seasonal variation remain unchanged as the climate warms, (3) warming is uniform across all months. Limitations include the fact that in reality, winters may warm faster than summers (changing the amplitude), and the warming rate may accelerate over time. A more sophisticated model might let a and d both depend on n.

SUMMARY

Lesson Summary

Summary

Periodic function modeling in AHL 3.8 centres on using sinusoidal functions of the form y = a sin(b(x − c)) + d or y = a cos(b(x − c)) + d to represent real-world repeating phenomena. The four key parameters each carry contextual meaning: the amplitude a measures the extent of oscillation, the b-value determines the period T = 2π/b, the phase shift c aligns the model with the timing of observed peaks, and the vertical shift d sets the midline equal to the average of the maximum and minimum values.

To build a model from data, follow the systematic six-step process: identify max and min, compute a and d, determine the period, calculate b, find the phase shift, and assemble the equation. Remember that sinusoidal models assume symmetric oscillation and a constant period — limitations you should be prepared to discuss in IB exam responses. These models connect forward to differentiation and integration of trigonometric functions and to sinusoidal regression on your GDC.