IVPs & Existence/Uniqueness — Initial Value Problems (IVPs)...

SECTION 1

Historical Context & Motivation

Differential equations describe how quantities change, and they show up everywhere — from the arc of a basketball to the spread of a virus. But solving a differential equation usually produces a whole family of curves, not just one answer. Early mathematicians realized they needed something extra — a starting condition — to pin down the single curve that matches a real-world situation. This idea became the initial value problem (IVP), one of the most important setups in all of applied mathematics.

1687

Newton's Principia

Isaac Newton formulated his laws of motion using what we now call differential equations. He implicitly used initial positions and velocities to predict planetary orbits — the earliest spirit of an IVP.

1760s

Euler's Systematic Methods

Leonhard Euler developed numerical step-by-step methods for solving differential equations from a given starting point, formalizing the concept that a single initial value could determine an entire solution curve.

1820s

Cauchy's Rigorous Foundations

Augustin-Louis Cauchy proved the first rigorous existence theorem, showing that under certain smoothness conditions a unique solution must exist near the starting point. This laid the groundwork for the modern existence and uniqueness theory.

1890

Picard–Lindelöf Theorem

Émile Picard and Ernst Lindelöf independently refined the conditions that guarantee both existence and uniqueness of solutions. Their theorem remains the standard reference taught in courses today.

The central question that drove centuries of work is deceptively simple: Given a rule for how something changes and a place to start, does exactly one future path exist? Understanding when the answer is "yes" — and when it might be "no" — is the focus of this lesson.

SECTION 2

Core Principles & Definitions

Before we dive into theorems, let's nail down the vocabulary. A differential equation is any equation that involves an unknown function and its derivatives. When we attach a specific starting value, we turn it into an initial value problem. The key ideas below form the foundation for everything that follows.

Ordinary Differential Equation (ODE)

An equation relating an unknown function y(t) to one or more of its derivatives with respect to a single independent variable t. Example: dy/dt = 3y.

Initial Condition

A statement that fixes the value of the unknown function at a specific point, such as y(0) = 5. This 'starting snapshot' anchors the solution.

General vs. Particular Solution

The general solution is the full family of curves (with an arbitrary constant C). The particular solution is the single member of that family picked out by the initial condition.

Existence

Asks: 'Does at least one solution exist that satisfies both the ODE and the initial condition?' If the function in the ODE behaves nicely enough, the answer is yes.

Uniqueness

Asks: 'Is there at most one solution?' Even when a solution exists, there might be more than one. Uniqueness requires an additional smoothness condition on the ODE.

✦ KEY TAKEAWAY

Think of a differential equation as a map that shows every possible road leading away from every point. The initial condition is like dropping a pin on that map — it tells you where your trip begins. Existence means there is at least one road out of that pin. Uniqueness means there is exactly one road — no confusing forks.

SECTION 3

Visual Explanation — Solution Families & the Role of the Initial Condition

The diagram below shows the general solution of dy/dt = y, which is y = Ce^t. Each curve corresponds to a different value of the constant C. Notice how the curves fan out — every one satisfies the same differential equation but starts at a different height. The red dot marks the initial condition y(0) = 2, which selects exactly one curve from the family.

The violet curves show members of the general solution y = Ce^t for various constants C. The cyan curve (C = 2) is the particular solution selected by the initial condition y(0) = 2, shown as the red dot.

Notice how every curve in the family satisfies the same differential equation dy/dt = y, but each starts at a different height on the y-axis. The initial condition acts like a filter: out of infinitely many possibilities, it selects exactly one particular solution. This is the geometric essence of an IVP — one point determines one path.

SECTION 4

Mathematical Framework

Let's formalize the pieces. A first-order IVP combines a differential equation with an initial condition into a single package. We then state the two theorems that tell us when solutions behave well.

INITIAL VALUE PROBLEM (FIRST ORDER)

dy/dt = f(t, y), y(t₀) = y₀

Here f(t, y) is a given function that defines the rate of change, t₀ is the starting time, and y₀ is the starting value of y.

Existence Theorem (Peano's Theorem)

If f(t, y) is continuous on some rectangle around the point (t₀, y₀), then there exists at least one solution y(t) defined on some interval containing t₀. In plain terms: if the rate-of-change function has no sudden jumps or holes near your starting point, you can always find a curve that fits.

EXISTENCE CONDITION

f(t, y) continuous near (t₀, y₀) ⟹ at least one solution exists

Continuity means no breaks, jumps, or division by zero in f near the starting point.

Uniqueness Condition (Picard–Lindelöf Theorem)

Existence alone doesn't rule out multiple solutions branching from the same starting point. To guarantee uniqueness, we need a stronger condition. The Picard–Lindelöf theorem says: if f is continuous and ∂f/∂y (the partial derivative of f with respect to y) is also continuous near (t₀, y₀), then the solution is the only one.

EXISTENCE + UNIQUENESS CONDITION

f and ∂f/∂y both continuous near (t₀, y₀) ⟹ exactly one solution

∂f/∂y is the partial derivative of f with respect to y. You can think of it as measuring how sensitive the rate of change is to small changes in y. If this sensitivity stays finite and well-behaved, solutions can't split apart.

🚀 Why Uniqueness Matters

Imagine you're building a simulation of a rocket's trajectory. If two different paths satisfy the same equation with the same launch data, you have no way to know which one the rocket will actually follow. Uniqueness guarantees your prediction is trustworthy.

SECTION 5

When Existence or Uniqueness Fails

The theorems above have conditions for a reason — not every differential equation plays nicely. The classic example of uniqueness failure is the IVP dy/dt = y^1/3 with y(0) = 0. Here f(t, y) = y^1/3 is continuous everywhere, so at least one solution exists. But ∂f/∂y = (1/3)y^−2/3 blows up at y = 0, so the uniqueness condition is not satisfied. Indeed, both y(t) = 0 and y(t) = (2t/3)^3/2 solve the same IVP — two different paths from the same starting point.

Both the cyan line (y = 0) and the red curve pass through the same initial point (0, 0) and both satisfy dy/dt = y^1/3. This is uniqueness failure in action.

This example is not just a mathematical curiosity. It highlights a critical practical lesson: when you set up a model, you should always check the hypotheses of the existence and uniqueness theorems. If ∂f/∂y has a blow-up at your starting point, your model may predict multiple futures — a sign that the equation needs to be refined or interpreted more carefully.

Checking existence and uniqueness conditions for various IVPs

Scenario	f continuous?	∂f/∂y continuous?	Conclusion
dy/dt = 2t, y(0) = 1	Yes	Yes (∂f/∂y = 0)	Unique solution exists
dy/dt = y^1/3, y(0) = 0	Yes	No (blows up at y = 0)	Exists but NOT unique
dy/dt = 1/t, y(0) = 3	No (undefined at t = 0)	N/A	Existence not guaranteed
dy/dt = 3y, y(1) = 5	Yes	Yes (∂f/∂y = 3)	Unique solution exists

SECTION 6

Worked Example

Let's walk through a complete IVP from start to finish: solve dy/dt = −2y with y(0) = 6. We'll verify the existence/uniqueness conditions, find the general solution, and then apply the initial condition.

Solve the IVP: dy/dt = −2y, y(0) = 6

Step 1 — Check Existence & Uniqueness ConditionsHere f(t, y) = −2y. This function is continuous everywhere. The partial derivative ∂f/∂y = −2, which is also continuous everywhere. Since both conditions of the Picard–Lindelöf theorem are satisfied near (0, 6), we are guaranteed a unique solution exists.

Both conditions met → unique solution guaranteed ✓

Step 2 — Separate VariablesRewrite the equation as dy/y = −2 dt. We have moved all y terms to one side and all t terms to the other. This technique is called separation of variables.

dy/y = −2 dt

Step 3 — Integrate Both SidesIntegrating the left side gives ln|y|. Integrating the right side gives −2t + C, where C is an arbitrary constant.

ln|y| = −2t + C

Step 4 — Solve for y (General Solution)Exponentiate both sides to undo the natural log: |y| = e^{−2t + C} = e^C · e^−2t. Letting A = ±e^C absorb the sign and constant, we get the general solution.

y = Ae^−2t

Step 5 — Apply the Initial ConditionSubstitute t = 0 and y = 6 into y = Ae^−2t: 6 = A · e⁰ = A · 1 = A. So A = 6.

y(t) = 6e^−2t — the unique particular solution

This solution describes exponential decay — the quantity starts at 6 and shrinks toward zero as t increases. Because we confirmed existence and uniqueness before solving, we know this is the only function that satisfies both the ODE and the initial condition.

SECTION 7

Strengths & Limitations of the Existence/Uniqueness Theorems

The existence and uniqueness theorems are powerful, but they have limits. Understanding both sides will help you apply them wisely and avoid common pitfalls.

Balancing what the theorems can and cannot do

Strengths	Limitations
Guarantee a unique solution exists before you spend time solving — like checking you have all recipe ingredients before you start cooking.	They are local results — they guarantee a solution near the starting point but don't always tell you how far it extends.
Apply to a huge class of differential equations (any first-order ODE with sufficiently smooth f).	They don't actually tell you what the solution is — only that it exists. You still need techniques like separation of variables to find it.
Help identify trouble spots (singularities) where the model may break down.	Conditions are sufficient but not necessary — a solution might still be unique even if ∂f/∂y isn't continuous, but you can't rely on the theorem to confirm it.
Provide a foundation for numerical methods (e.g., Euler's method) by assuring the target solution is well-defined.	They don't cover systems with discontinuous forcing functions or certain boundary value problems without extensions.

✦ KEY TAKEAWAY

Think of the existence/uniqueness theorems as a safety checklist before a flight. They don't fly the plane for you (they don't solve the equation), but they confirm the runway is clear — meaning a well-defined solution path is waiting. When the checklist fails, you know to investigate further before trusting any answer.

SECTION 8

Connection to Advanced Topics

The ideas in this lesson are just the beginning. As you progress through differential equations, the existence/uniqueness framework generalizes to more complex settings. Here's a glimpse of where these ideas lead.

From introductory IVPs to advanced differential equations

This Lesson (Intro)	Advanced Extension
First-order IVPs: dy/dt = f(t, y), y(t₀) = y₀	Higher-order IVPs: y'' + p(t)y' + q(t)y = g(t) with y(t₀) and y'(t₀) given
Scalar equations (y is a single unknown)	Systems of equations (vectors of unknowns, e.g., position and velocity together)
Local existence (solution near t₀)	Global existence and blow-up analysis (does the solution last forever or go to infinity in finite time?)
Exact analytical solutions	Numerical methods (Euler, Runge-Kutta) for equations that can't be solved by hand
Initial value problems (conditions at one point)	Boundary value problems (conditions at two or more points — used in engineering and physics)

In every one of these extensions, the same fundamental question reappears: does a solution exist, and is it unique? The Picard–Lindelöf theorem you learned today is the template for all of them. Mastering it now gives you a conceptual passport to every advanced topic on this list.

SECTION 9

Practice Problems

PROBLEM 1 — CONCEPTUAL

In your own words, explain the difference between the general solution of a differential equation and a particular solution. What role does the initial condition play in connecting the two?

PROBLEM 2 — BASIC CALCULATION

Consider the IVP: dy/dt = 5y, y(0) = 3. Verify that the existence and uniqueness conditions are satisfied, then find the particular solution.

PROBLEM 3 — INTERMEDIATE

For the IVP dy/dt = t² + y², y(1) = 0, determine whether the existence and uniqueness theorem guarantees a unique solution near t = 1. Explain your reasoning. (You do not need to solve the equation.)

PROBLEM 4 — APPLIED

A tank initially contains 50 liters of a salt solution. Fresh water flows in at 2 liters per minute, and the well-mixed solution drains at 2 liters per minute. If A(t) is the amount of salt (in grams) at time t, the model is dA/dt = −(2/50)A = −0.04A with A(0) = 100 grams. Show that the existence/uniqueness conditions hold, solve the IVP, and find how much salt remains after 25 minutes.

PROBLEM 5 — CRITICAL THINKING

Consider dy/dt = √|y| with y(0) = 0. (a) Is f(t, y) = √|y| continuous at (0, 0)? (b) Is ∂f/∂y continuous at (0, 0)? (c) Can you find two different solutions that both satisfy this IVP? What does this tell you about the relationship between the continuity of f and the uniqueness of solutions?

SUMMARY

Lesson Summary

An initial value problem (IVP) pairs a differential equation dy/dt = f(t, y) with an initial condition y(t₀) = y₀. The general solution of the ODE is a family of curves parameterized by an arbitrary constant. The initial condition selects exactly one particular solution from that family. Whether such a solution actually exists and whether it is the only one depends on the behavior of f near the starting point.

The existence theorem (Peano) states that if f is continuous near (t₀, y₀), at least one solution exists. The Picard–Lindelöf theorem adds that if ∂f/∂y is also continuous, the solution is unique. When the uniqueness condition fails — as with dy/dt = y^1/3 at y = 0 — multiple solutions can emerge from the same starting point. Always check both conditions before trusting that your IVP has a single, well-defined answer.