The language that nature uses to describe change, motion, and the dynamics of everything around us
Look around you. A cup of coffee cooling on your desk. Your phone battery draining. A bridge swaying in the wind. A satellite orbiting Earth. The spread of a virus through a population. Everything that changes follows rules — and those rules are written in the language of differential equations.
A differential equation is simply a statement about how things change. When you write:
you are saying: "the rate at which \(y\) changes depends on where it is right now." That single idea — that the future depends on the present — is the foundation of virtually all mathematical modeling in engineering, physics, biology, economics, and beyond.
Differential equations are one of the tools through which we understand and appreciate the remarkable precision of the laws governing creation. Studying them is an act of reflection on the order and beauty embedded in the universe.
Here is the central insight that makes differential equations so powerful:
In most real-world systems, we don't know the solution directly, but we do know the rules of change. A differential equation encodes those rules. Solving the equation then gives us the complete behavior of the system — past, present, and future.
Think of it this way:
You drop a ball. You don't need to track its position at every instant. Instead, you know one rule: gravity pulls it downward with acceleration \(g = 9.8\) m/s². That rule is a differential equation:
From this single equation plus initial conditions (where you dropped it, how fast), you can predict the ball's position at any future time. That is the power of differential equations: local rules produce global predictions.
Before we start solving differential equations, we need a precise vocabulary to describe them. Just as a doctor classifies diseases before prescribing treatment, we classify DEs before choosing a solution method. Each classification tells us something about the equation's difficulty and which tools to use.
The first and most fundamental distinction: does the unknown function depend on one variable or several?
Unknown function depends on one variable
Here \(y = y(t)\) depends only on time \(t\).
Example: Temperature of coffee vs time
Unknown function depends on two or more variables
Here \(u = u(x, y)\) depends on two spatial variables.
Example: Heat distribution in a metal plate
In this course (Math 4), we focus entirely on ODEs. PDEs are studied in more advanced courses and require different techniques (Fourier series, separation of variables in multiple dimensions, etc.).
The order of a differential equation is the order of the highest derivative that appears in the equation.
This is simple to determine — just find the highest derivative:
| Equation | Highest Derivative | Order |
|---|---|---|
| \(\dfrac{dy}{dx} + 2y = x\) | \(y'\) (first derivative) | 1st order |
| \(\dfrac{d^2y}{dx^2} + 3\dfrac{dy}{dx} + y = 0\) | \(y''\) (second derivative) | 2nd order |
| \(y''' + y' = e^x\) | \(y'''\) (third derivative) | 3rd order |
| \(\dfrac{d^4y}{dx^4} = \sin(x)\) | \(y^{(4)}\) (fourth derivative) | 4th order |
The order determines how many initial conditions you need for a unique solution. A first-order ODE needs 1 condition (e.g., \(y(0) = 5\)), a second-order needs 2 (e.g., \(y(0) = 1,\; y'(0) = 0\)), and in general an \(n\)th-order ODE needs \(n\) initial conditions.
The degree of a differential equation is the power (exponent) of the highest-order derivative, after the equation has been cleared of fractions and radicals involving derivatives.
| Equation | Analysis | Order & Degree |
|---|---|---|
| \(\dfrac{dy}{dx} + y = x^2\) | \(y'\) appears to the 1st power | Order 1, Degree 1 |
| \(\left(\dfrac{dy}{dx}\right)^3 + y = 0\) | \(y'\) is raised to the 3rd power | Order 1, Degree 3 |
| \(\left(\dfrac{d^2y}{dx^2}\right)^2 + \dfrac{dy}{dx} = 1\) | \(y''\) is squared | Order 2, Degree 2 |
| \(\dfrac{d^2y}{dx^2} = \sqrt{1 + \left(\dfrac{dy}{dx}\right)^2}\) | Square both sides first: \((y'')^2 = 1 + (y')^2\). Now \(y''\) has power 2. | Order 2, Degree 2 |
| \(\dfrac{dy}{dx} = \sin(y')\) | Cannot be written as polynomial in \(y'\) | Order 1, Degree undefined |
Degree is not the same as order! Also, if the derivatives appear inside transcendental functions (sin, cos, exp, log), the degree is not defined. The equation must be expressible as a polynomial in the highest derivative for degree to be meaningful.
An ODE is linear if it can be written in the form:
The key requirements are:
• The unknown \(y\) and all its derivatives appear to the first power only
• There are no products of \(y\) with its derivatives (no \(y \cdot y'\))
• The coefficients \(a_n(x), \ldots, a_0(x)\) and \(g(x)\) depend only on \(x\), not on \(y\)
An equation that is not linear is called nonlinear. Let's see examples:
| Equation | Linear or Nonlinear? | Why? |
|---|---|---|
| \(y' + 2y = x^3\) | Linear | \(y\) and \(y'\) both to the 1st power, coefficients depend only on \(x\) |
| \(y'' + 9y = 0\) | Linear | Standard 2nd-order linear (simple harmonic oscillator) |
| \(x^2y'' + xy' + y = \sin(x)\) | Linear | Variable coefficients \(x^2, x\) are fine — they depend on \(x\), not \(y\) |
| \(y' + y^2 = 0\) | Nonlinear | \(y\) appears squared — \(y^2\) violates linearity |
| \(y \cdot y' = x\) | Nonlinear | Product of \(y\) and \(y'\) — not allowed in linear equations |
| \(y'' + \sin(y) = 0\) | Nonlinear | \(\sin(y)\) is a nonlinear function of \(y\) (the pendulum equation!) |
| \(y' = e^y\) | Nonlinear | \(e^y\) is nonlinear in \(y\) |
Linear equations have a beautiful property: the superposition principle. If \(y_1\) and \(y_2\) are solutions to a linear homogeneous equation, then \(c_1 y_1 + c_2 y_2\) is also a solution. This means we can build complex solutions from simple ones. Nonlinear equations don't have this luxury — they are fundamentally harder and often require special techniques or numerical methods.
A linear ODE is homogeneous if \(g(x) = 0\) (the right-hand side is zero):
If \(g(x) \neq 0\), the equation is non-homogeneous:
No external forcing. Solutions describe natural behavior (free vibrations, unforced response).
External forcing present. Solutions include forced response (driven oscillations, input-output).
The type of conditions given alongside the DE determines whether you have an IVP or BVP:
All conditions given at the same point (here, \(x = 0\)). Think of knowing the starting position and velocity.
Conditions given at different points (here, two endpoints). Think of knowing the temperature at both ends of a rod.
We primarily study IVPs for first- and second-order ODEs. BVPs appear in more advanced topics like heat conduction and structural analysis.
When you encounter a differential equation, ask these questions in order:
| Question | Options | How to Tell |
|---|---|---|
| ODE or PDE? | Ordinary / Partial | Does the unknown depend on one variable or several? |
| What order? | 1st, 2nd, 3rd, … | Find the highest derivative present |
| What degree? | 1, 2, 3, … or undefined | Power of the highest derivative (after clearing radicals) |
| Linear or nonlinear? | Linear / Nonlinear | Is \(y\) (and its derivatives) to the 1st power only, with no products? |
| Homogeneous? | Homogeneous / Non-homogeneous | Is the right-hand side zero? (Only applies to linear equations) |
| IVP or BVP? | Initial / Boundary | Are all conditions at one point, or at different points? |
Try classifying each equation before reading the answer:
\(\displaystyle \frac{d^2y}{dx^2} + 3\frac{dy}{dx} - 2y = e^x\) → ODE, 2nd order, degree 1, linear, non-homogeneous
\(\displaystyle \left(\frac{dy}{dx}\right)^2 + y = x\) → ODE, 1st order, degree 2, nonlinear
\(\displaystyle y''' + y \cdot y'' = 0\) → ODE, 3rd order, degree 1, nonlinear (because of the \(y \cdot y''\) product)
\(\displaystyle \frac{\partial^2 u}{\partial t^2} = c^2\frac{\partial^2 u}{\partial x^2}\) → PDE, 2nd order, degree 1, linear, homogeneous (the wave equation!)
Let's walk through several examples from different fields — each one a real system you encounter in life or engineering, modeled by a simple first-order DE.
You pour a cup of coffee at 90°C. The room is 22°C. How does the temperature change over time?
Newton's Law of Cooling says: the rate of cooling is proportional to the temperature difference between the coffee and the room.
Here \(k > 0\) is a constant that depends on the cup material, airflow, etc. This is a separable and linear first-order ODE!
Solution: \(T(t) = 22 + 68\,e^{-kt}\)
Your phone battery at 100% loses charge at a rate roughly proportional to the number of active processes, which in turn depends on the current charge level (as low-power mode kicks in).
A simplified model:
where \(B(t)\) is the battery percentage and \(\alpha\) depends on usage.
Solution: \(B(t) = 100\,e^{-\alpha t}\)
If a population of bacteria doubles every hour, how many are there after \(t\) hours?
The growth rate is proportional to the current population:
With \(P(0) = P_0\), the solution is \(P(t) = P_0 e^{rt}\).
But resources are limited! With a carrying capacity \(K\), we get the logistic equation:
Connect a 12V battery to a resistor (\(R = 1\,k\Omega\)) and capacitor (\(C = 1\,mF\)) in series. What happens to the voltage across the capacitor?
Kirchhoff's voltage law gives:
Solution: \(v(t) = 12(1 - e^{-t})\)
You take a 500mg dose of medicine. Your body eliminates it at a rate proportional to the current amount:
The half-life is the time for half the drug to be eliminated: \(t_{1/2} = \frac{\ln 2}{\lambda}\).
If the half-life is 4 hours: \(A(t) = 500 \cdot 2^{-t/4}\)
A 200-liter tank has pure water. Brine (0.1 kg/L of salt) flows in at 5 L/min, and the well-mixed solution flows out at 5 L/min. How much salt is in the tank at time \(t\)?
Solution: \(A(t) = 20(1 - e^{-t/40})\) kg
The simplest DE in the world — \(\dfrac{dy}{dt} = ky\) — produces exponential behavior. Adjust \(k\) to see how the sign determines growth vs decay:
Set \(k > 0\) for exponential growth (populations, compound interest). Set \(k < 0\) for decay (radioactivity, cooling). Set \(k = 0\) and the system is static — nothing changes. The sign of one number determines the entire qualitative behavior!
Differential equations are not just abstract math — they are one of humanity's greatest intellectual achievements, born from the desire to understand the physical world.
Every time you use DEs to solve an engineering problem, you follow this cycle:
This course teaches you step 2 (Solve) — the mathematical tools to turn differential equations into solutions. But always remember: the math serves the modeling. Every technique you learn here is a tool for understanding the real world.
The pursuit of mathematical knowledge is a noble endeavor. Every equation you solve, every concept you master, brings you closer to understanding the magnificent precision with which this universe operates. Approach this course with curiosity, patience, and the intention to benefit yourself and others with what you learn.
Start with Chapter 1: First-Order Differential Equations, or return to the Course Home to see the full roadmap.