Scribe logoScribeNotes & practice
Sign inStart free
  1. Notes
  2. ›IB Mathematics: Applications and Interpretation
  3. ›5 Calculus
  4. ›5.17 Phase portraits

IB Mathematics: Applications and Interpretation · Topic 5 · Calculus

5.17 Phase portraits

Identify equilibrium points of a linear system, classify them using eigenvalues, and describe what each phase portrait shows about long-term behaviour.

  • 4 min read
  • 5 parts
  • 10 questions
  • Draft
  • HL only
Start a Quick 5 on this topic→Jump to the 10 questions
1/5A pair of rate equations describes two linked quantities
  1. 1A pair of rate equations describes two linked quantities
  2. 2Equilibrium points are where both rates equal zero
  3. 3The eigenvalues of the coefficients decide the pattern
  4. 4Solving the quadratic classifies any linear equilibrium
  5. 5Trajectories show how a system behaves in the long run
  6. Check your recall
  7. Practice: 10 questions
1

Part 1 of 5: A pair of rate equations describes two linked quantities

Think of two populations that affect each other, such as a predator and its prey. Each population changes at a rate that depends on both. This gives two linked differential equations, one for xxx and one for yyy.

In the linear case the system is dxdt=ax+by\frac{dx}{dt}=ax+bydtdx​=ax+by and dydt=cx+dy\frac{dy}{dt}=cx+dydtdy​=cx+dy. The letters aaa, bbb, ccc and ddd are constants. Time ttt does not appear on the right-hand sides.

Plot xxx along one axis and yyy along the other. This plane is the phase plane. As ttt increases, the point (x,y)(x,y)(x,y) traces a curve called a trajectory. Arrows on the curve show the direction of travel.

Key terms · tap one for its definition

phase plane

The phase plane is the plane whose axes are the two dependent variables xxx and yyy of a system, with time as a hidden parameter.

trajectory

A trajectory is the curve traced by the point (x,y)(x,y)(x,y) in the phase plane as ttt increases.

Part 1 done

2

Part 2 of 5: Equilibrium points are where both rates equal zero

Now ask where a trajectory would simply stop. That happens when neither quantity is changing. A ball resting at the bottom of a bowl is a good picture.

A point where dxdt=0\frac{dx}{dt}=0dtdx​=0 and dydt=0\frac{dy}{dt}=0dtdy​=0 together is an equilibrium point. A system that starts there stays there for ever. To find it, set both right-hand sides to zero and solve the two equations at once.

Worked example

Finding an equilibrium point

Find the equilibrium point of dxdt=8−2x\frac{dx}{dt}=8-2xdtdx​=8−2x and dydt=x−4y\frac{dy}{dt}=x-4ydtdy​=x−4y.
  1. 1
    Set both rates to zero: 8−2x=08-2x=08−2x=0 and x−4y=0x-4y=0x−4y=0.
  2. 2
    Solve the first equation: x=4x=4x=4.
  3. 3
    Substitute into the second equation: 4−4y=04-4y=04−4y=0, so y=1y=1y=1.

Answer: The equilibrium point is (4,1)(4,1)(4,1).

Common mistake

Setting only one rate to zero gives a whole curve of points. An equilibrium needs both rates to be zero at the same point.

Key term

equilibrium point

An equilibrium point is a point in the phase plane where dxdt=0\frac{dx}{dt}=0dtdx​=0 and dydt=0\frac{dy}{dt}=0dtdy​=0 at the same time.

Part 2 done

Quick check

One question on the part you just read

Easy

Identify the condition that defines an equilibrium point of a system with rates dxdt\frac{dx}{dt}dtdx​ and dydt\frac{dy}{dt}dtdy​.
Show answer and explanation

Answer: C · Worked explanation

An equilibrium point needs both rates to be zero together, so neither quantity changes. Requiring only one rate to be zero gives a whole curve of points where just one quantity is momentarily still. Equal rates only mean that xxx and yyy change at the same speed, not that they stop.
3

Part 3 of 5: The eigenvalues of the coefficients decide the pattern

Knowing where the equilibrium is does not say what happens near it. Trajectories may approach it, leave it or circle it.

For the linear system, the pattern depends on a number called an eigenvalue, written λ\lambdaλ. A system has two eigenvalues. They solve λ2−(a+d)λ+(ad−bc)=0\lambda^2-(a+d)\lambda+(ad-bc)=0λ2−(a+d)λ+(ad−bc)=0. The two roots are either real or a complex pair.

An equilibrium point is stable when trajectories nearby are drawn back to it. Otherwise it is unstable, or neutral when trajectories circle it. Real parts of the eigenvalues control growth or decay. Imaginary parts control rotation.

Patterns near an equilibrium point

EigenvaluesNameBehaviour near the point
Both real, negativeStable nodeTrajectories approach the point directly
Both real, positiveUnstable nodeTrajectories leave the point directly
Real, opposite signsSaddleSome trajectories approach, then most leave
Complex, negative real partStable spiralTrajectories spiral in
Complex, positive real partUnstable spiralTrajectories spiral out
Purely imaginaryCentreClosed loops around the point
Eigenvalues
Both real, negative
Name
Stable node
Behaviour near the point
Trajectories approach the point directly
Eigenvalues
Both real, positive
Name
Unstable node
Behaviour near the point
Trajectories leave the point directly
Eigenvalues
Real, opposite signs
Name
Saddle
Behaviour near the point
Some trajectories approach, then most leave
Eigenvalues
Complex, negative real part
Name
Stable spiral
Behaviour near the point
Trajectories spiral in
Eigenvalues
Complex, positive real part
Name
Unstable spiral
Behaviour near the point
Trajectories spiral out
Eigenvalues
Purely imaginary
Name
Centre
Behaviour near the point
Closed loops around the point
Figure 1. On the left every trajectory runs into the equilibrium point. On the right trajectories come in along one axis and leave along the other.

Key terms · tap one for its definition

eigenvalue

An eigenvalue is a root λ\lambdaλ of the equation λ2−(a+d)λ+(ad−bc)=0\lambda^2-(a+d)\lambda+(ad-bc)=0λ2−(a+d)λ+(ad−bc)=0 built from the coefficients of a linear system.

stable

A stable equilibrium point is one that every trajectory starting close enough to it approaches as ttt increases without limit.

Part 3 done

4

Part 4 of 5: Solving the quadratic classifies any linear equilibrium

Now put the steps together. The system from the earlier example has its equilibrium at (4,1)(4,1)(4,1). The method below classifies it.

  • Find the equilibrium point by setting both rates to zero.
  • Read off aaa, bbb, ccc and ddd from the coefficients.
  • Solve λ2−(a+d)λ+(ad−bc)=0\lambda^2-(a+d)\lambda+(ad-bc)=0λ2−(a+d)λ+(ad−bc)=0.
  • Match the eigenvalues to the table.

When the equilibrium is not the origin, the same coefficients still decide the pattern. Shifting the axes to the equilibrium does not change them.

Worked example

Classifying an equilibrium

Classify the equilibrium point (4,1)(4,1)(4,1) of dxdt=8−2x\frac{dx}{dt}=8-2xdtdx​=8−2x and dydt=x−4y\frac{dy}{dt}=x-4ydtdy​=x−4y.
  1. 1
    Read the coefficients: a=−2a=-2a=−2, b=0b=0b=0, c=1c=1c=1 and d=−4d=-4d=−4.
  2. 2
    Find a+d=−6a+d=-6a+d=−6 and ad−bc=(−2)(−4)−(0)(1)=8ad-bc=(-2)(-4)-(0)(1)=8ad−bc=(−2)(−4)−(0)(1)=8.
  3. 3
    Form the quadratic: λ2+6λ+8=0\lambda^2+6\lambda+8=0λ2+6λ+8=0.
  4. 4
    Factorise: (λ+2)(λ+4)=0(\lambda+2)(\lambda+4)=0(λ+2)(λ+4)=0, so λ=−2\lambda=-2λ=−2 or λ=−4\lambda=-4λ=−4.
  5. 5
    Both eigenvalues are real and negative.

Answer: The equilibrium point is a stable node.

Exam tip

Write the eigenvalues, then name the pattern, then state the stability. Each statement can earn its own mark.

Part 4 done

Quick check

One question on the part you just read

Easy

State what the arrows drawn on a trajectory show.
Show answer and explanation

Answer: A · Worked explanation

The arrows show the direction of travel as time passes. Both xxx and yyy can increase or decrease along one trajectory, so neither variable fixes the arrow. Pointing towards the nearest equilibrium point is wrong, because trajectories near an unstable point move away from it.
5

Part 5 of 5: Trajectories show how a system behaves in the long run

Once an equilibrium is classified, the portrait describes what happens over time. Read the arrows first, because they show where each trajectory is heading.

  • Stable node: both quantities settle at their equilibrium values without oscillating.
  • Stable spiral: both quantities oscillate with shrinking amplitude while settling.
  • Unstable node or spiral: trajectories move away, so the quantities leave the equilibrium values.
  • Saddle: most trajectories approach, then leave along a different direction.
  • Centre: both quantities repeat the same cycle for ever.

Common mistake

A saddle is unstable even though some trajectories approach it. Only trajectories along one special direction do.

Go deeperWhere the eigenvalues come from

Solutions of a linear system are built from terms eλte^{\lambda t}eλt. A negative λ\lambdaλ makes a term decay, and a positive λ\lambdaλ makes it grow.

For complex eigenvalues α±βi\alpha\pm\beta\mathrm{i}α±βi, the solutions contain eαtcos⁡βte^{\alpha t}\cos\beta teαtcosβt and eαtsin⁡βte^{\alpha t}\sin\beta teαtsinβt. The real part α\alphaα controls growth or decay, and β\betaβ controls rotation.

For nonlinear systems, a portrait near an equilibrium usually resembles the linear case, so the same classification applies close to that point.

Note read

Check your recall[15 marks]

6 written questions. Answer each one in your head or on paper, then open the mark scheme.

  1. 1.Define[2 marks]

    Define an equilibrium point of a system with rates dxdt\frac{dx}{dt}dtdx​ and dydt\frac{dy}{dt}dtdy​.

    Show mark schemeHide mark scheme

    Model answer

    An equilibrium point is a point where dxdt=0\frac{dx}{dt}=0dtdx​=0 and dydt=0\frac{dy}{dt}=0dtdy​=0 at the same time. A system that starts there does not change.

    • 1 mark for stating that both dxdt\frac{dx}{dt}dtdx​ and dydt\frac{dy}{dt}dtdy​ are zero
    • 1 mark for stating that this happens at the same point, so a system starting there stays there
  2. 2.Determine[2 marks]

    Determine the equilibrium point of dxdt=12−3x\frac{dx}{dt}=12-3xdtdx​=12−3x and dydt=x−2y\frac{dy}{dt}=x-2ydtdy​=x−2y.

    Show mark schemeHide mark scheme

    Model answer

    Setting 12−3x=012-3x=012−3x=0 gives x=4x=4x=4. Then 4−2y=04-2y=04−2y=0 gives y=2y=2y=2. The equilibrium point is (4,2)(4,2)(4,2).

    • 1 mark for solving 12−3x=012-3x=012−3x=0 to obtain x=4x=4x=4
    • 1 mark for substituting to obtain y=2y=2y=2
  3. 3.Calculate[4 marks]

    Calculate the eigenvalues of dxdt=x+2y\frac{dx}{dt}=x+2ydtdx​=x+2y and dydt=3x+2y\frac{dy}{dt}=3x+2ydtdy​=3x+2y, and hence classify the equilibrium point at the origin.

    Show mark schemeHide mark scheme

    Model answer

    Here a+d=3a+d=3a+d=3 and ad−bc=2−6=−4ad-bc=2-6=-4ad−bc=2−6=−4. So λ2−3λ−4=0\lambda^2-3\lambda-4=0λ2−3λ−4=0, giving λ=4\lambda=4λ=4 and λ=−1\lambda=-1λ=−1. The eigenvalues are real with opposite signs, so the origin is a saddle and is unstable.

    • 1 mark for finding a+d=3a+d=3a+d=3 and ad−bc=−4ad-bc=-4ad−bc=−4
    • 1 mark for forming λ2−3λ−4=0\lambda^2-3\lambda-4=0λ2−3λ−4=0
    • 1 mark for obtaining λ=4\lambda=4λ=4 and λ=−1\lambda=-1λ=−1
    • 1 mark for stating that opposite signs give a saddle, which is unstable
  4. 4.Describe[2 marks]

    Describe the behaviour of trajectories near a stable spiral.

    Show mark schemeHide mark scheme

    Model answer

    Trajectories spiral inwards towards the equilibrium point. Both quantities oscillate with shrinking amplitude as they settle.

    • 1 mark for stating that trajectories spiral in towards the equilibrium point
    • 1 mark for stating that the quantities oscillate with decreasing amplitude
  5. 5.Distinguish[3 marks]

    Distinguish between a saddle and a stable node.

    Show mark schemeHide mark scheme

    Model answer

    A stable node has two negative real eigenvalues, and every nearby trajectory approaches the point. A saddle has real eigenvalues of opposite signs. Most trajectories leave, so it is unstable.

    • 1 mark for stating that a stable node has two negative real eigenvalues
    • 1 mark for stating that a saddle has real eigenvalues of opposite signs
    • 1 mark for stating that a node attracts nearby trajectories whereas a saddle is unstable because most leave
  6. 6.Explain[2 marks]

    Explain why a centre is not a stable equilibrium point.

    Show mark schemeHide mark scheme

    Model answer

    A centre has purely imaginary eigenvalues, so trajectories are closed loops. They circle the point and do not approach it as ttt increases.

    • 1 mark for stating that the eigenvalues are purely imaginary and trajectories are closed loops
    • 1 mark for explaining that trajectories do not approach the point as ttt increases

Practice this topic

10 quick-fire multiple-choice questions, easy to hard. Tap an answer to see if you're right and the worked explanation. (The written questions with mark schemes are under “Check your recall” above.)

0 right0 of 10 answered

Includes the 2 quick checks in the note above.

3.

Easy·Deduce[1 mark]

Deduce the pattern near an equilibrium point when the two eigenvalues are the real numbers -2 and -5.
Show answer and explanation

Answer: D · Worked explanation

Two real eigenvalues that are both negative give a stable node, so trajectories approach the point directly. A saddle would need one positive and one negative eigenvalue. A spiral would need a complex pair, and these eigenvalues are real. The signs of the eigenvalues are what separate these cases.
4.

Medium·Determine[2 marks]

Determine the equilibrium point of dxdt=3x−12\frac{dx}{dt}=3x-12dtdx​=3x−12 and dydt=2x−y\frac{dy}{dt}=2x-ydtdy​=2x−y.
Show answer and explanation

Answer: B · Worked explanation

Setting 3x−12=03x-12=03x−12=0 gives x=4x=4x=4. Then 2x−y=02x-y=02x−y=0 gives y=2x=8y=2x=8y=2x=8. The point (4,2)(4,2)(4,2) comes from dividing by two instead of multiplying, and (8,4)(8,4)(8,4) swaps the coordinates. Substituting the result into both equations confirms that both rates are zero, which is a useful check in any exam answer.
5.

Medium·Determine[2 marks]

Determine the pattern at the origin for dxdt=−3x+y\frac{dx}{dt}=-3x+ydtdx​=−3x+y and dydt=x−3y\frac{dy}{dt}=x-3ydtdy​=x−3y.
Show answer and explanation

Answer: A · Worked explanation

Here a+d=−6a+d=-6a+d=−6 and ad−bc=9−1=8ad-bc=9-1=8ad−bc=9−1=8. So λ2+6λ+8=0\lambda^2+6\lambda+8=0λ2+6λ+8=0, giving λ=−2\lambda=-2λ=−2 and λ=−4\lambda=-4λ=−4. Both are real and negative. A saddle would need ad−bc<0ad-bc<0ad−bc<0, but this value is +8+8+8. Both eigenvalues being negative means every trajectory approaches the origin. An unstable node would need both to be positive.
6.

Medium·Calculate[2 marks]

Calculate the eigenvalues of dxdt=−x+2y\frac{dx}{dt}=-x+2ydtdx​=−x+2y and dydt=−2x−y\frac{dy}{dt}=-2x-ydtdy​=−2x−y, and name the pattern.
Show answer and explanation

Answer: C · Worked explanation

Here a+d=−2a+d=-2a+d=−2 and ad−bc=1+4=5ad-bc=1+4=5ad−bc=1+4=5. So λ2+2λ+5=0\lambda^2+2\lambda+5=0λ2+2λ+5=0, giving λ=−1±2i\lambda=-1\pm2\mathrm{i}λ=−1±2i. The negative real part means trajectories spiral in. The values -3 and 1 come from adding bcbcbc instead of subtracting it. A centre would need a real part of zero, but here it is -1.
7.

Medium·Predict[1 mark]

A model of two populations has a stable spiral at its equilibrium point. Predict the long-term behaviour.
Show answer and explanation

Answer: B · Worked explanation

A stable spiral draws trajectories in, so the quantities settle. The spiral shape means each quantity rises and falls on the way. Constant amplitude would describe a centre, not a stable spiral. The spiral is the usual picture of damped oscillation in a two-quantity model.
8.

Hard·Identify[2 marks]

Identify the correct statement about a saddle.
Show answer and explanation

Answer: D · Worked explanation

A saddle has one negative and one positive eigenvalue. Trajectories along one special direction approach it, but all others eventually leave, so it is unstable. Having some trajectories approach is tempting evidence of stability, but stability needs every nearby trajectory to approach. Only one special direction leads in, and every other path eventually turns away.
9.

Hard·Determine[2 marks]

Determine the equilibrium point and its pattern for dxdt=x−2y+3\frac{dx}{dt}=x-2y+3dtdx​=x−2y+3 and dydt=2x+y−4\frac{dy}{dt}=2x+y-4dtdy​=2x+y−4.
Show answer and explanation

Answer: A · Worked explanation

Setting both rates to zero gives x−2y=−3x-2y=-3x−2y=−3 and 2x+y=42x+y=42x+y=4, so y=2y=2y=2 and x=1x=1x=1. Here a+d=2a+d=2a+d=2 and ad−bc=1+4=5ad-bc=1+4=5ad−bc=1+4=5, so λ=1±2i\lambda=1\pm2\mathrm{i}λ=1±2i. The real part is positive, so trajectories spiral out. A stable spiral would need a negative real part.
10.

Hard·Determine[2 marks]

For dxdt=ky\frac{dx}{dt}=kydtdx​=ky and dydt=−x\frac{dy}{dt}=-xdtdy​=−x, determine the value of kkk that makes the origin a centre.
Show answer and explanation

Answer: D · Worked explanation

Here a=d=0a=d=0a=d=0, b=kb=kb=k and c=−1c=-1c=−1, so λ2+k=0\lambda^2+k=0λ2+k=0. For k=4k=4k=4, λ=±2i\lambda=\pm2\mathrm{i}λ=±2i, which is purely imaginary, so the origin is a centre. A negative kkk gives real eigenvalues of opposite signs, which is a saddle. The value k=0k=0k=0 gives no loops. The sign of kkk therefore decides between a saddle and a centre.

Keep going

Next topic5.18 Second-order differential equationsPrevious topic5.16 Euler's method

More in Topic 5 · Calculus

  1. 5.10Second derivativesNotes · 10 questions· HL only
  2. 5.11Further integrationNotes · 10 questions· HL only
  3. 5.12Areas and volumes of revolutionNotes · 10 questions· HL only
  4. 5.13KinematicsNotes · 10 questions· HL only
  5. 5.14Setting up differential equationsNotes · 10 questions· HL only
  6. 5.15Slope fieldsNotes · 10 questions· HL only
  7. 5.16Euler's methodNotes · 10 questions· HL only
  8. 5.17Phase portraitsNotes · 10 questions· HL only
  9. 5.18Second-order differential equationsNotes · 10 questions· HL only

All IB Mathematics: Applications and Interpretation topics →

Draft: written by Scribe and checked automatically, not yet by a subject reviewer.

Practice questions are original, written by Scribe and checked automatically for correctness.

Topic codes follow the official numbering. Not endorsed by IB Diploma Programme or any exam board.

On this page

0/5 read

  1. 1A pair of rate equations describes two linked quantities
  2. 2Equilibrium points are where both rates equal zero
  3. 3The eigenvalues of the coefficients decide the pattern
  4. 4Solving the quadratic classifies any linear equilibrium
  5. 5Trajectories show how a system behaves in the long run
  6. Check your recall
  7. Practice: 10 questions
All coursesScribeHow it worksPricingPrivacyTerms

Notes and questions are written by Scribe in our own words. Topic codes follow the official course numbering so you can find your place. Not produced or endorsed by the IB, College Board, AQA, OCR, Pearson Edexcel or any exam board.

© 2026 Scribe