evan's notes
cal 3discretelinear

On this page

  • Logistic population growth
  • Solving the logistic equation
  • A falling object with linear drag

Ordinary Differential Equations

§3 Logistic Growth and Falling Objects

Evan Luo · Sep 11, 2026

Ordinary Differential Equations

§3 Logistic Growth and Falling Objects

Evan LuoToday

3 min read

Two common models lead to first-order differential equations with stable equilibrium values. A logistic population approaches its carrying capacity, while a falling object's velocity approaches its terminal velocity.

Logistic population growth

Let N(t)N(t)N(t) be a population at time ttt. A population-dependent growth model has the form

dNdt=R(N)N,\frac{dN}{dt}=R(N)N, dtdN​=R(N)N,

where R(N)R(N)R(N) is the per-capita growth rate. To model limited resources, introduce a carrying capacity K>0K>0K>0. At zero population, the per-capita rate should equal the positive constant rrr; at capacity, it should vanish:

R(0)=r>0,R(K)=0.R(0)=r>0, \qquad R(K)=0. R(0)=r>0,R(K)=0.

It should also have the sign pattern

R(N)>0when 0<N<K,R(N)>0\quad\text{when }0<N<K, R(N)>0when 0<N<K,

so the population grows below capacity, and

R(N)<0when N>K,R(N)<0\quad\text{when }N>K, R(N)<0when N>K,

so it decreases above capacity. At capacity, R(K)=0R(K)=0R(K)=0.

The simplest function with this sign pattern is the decreasing line

R(N)=r(1−NK)=rK(K−N),r>0.R(N)=r\left(1-\frac{N}{K}\right) =\frac{r}{K}(K-N), \qquad r>0. R(N)=r(1−KN​)=Kr​(K−N),r>0.

Substituting it into the population model gives the logistic equation:

dNdt=rK(K−N)N.\boxed{\frac{dN}{dt}=\frac{r}{K}(K-N)N}. dtdN​=Kr​(K−N)N​.

It is first order, nonlinear because of the N2N^2N2 term, and separable.

Per-capita growth rate

Logistic per-capita growth rateR of N decreases linearly from r at population zero, crosses zero at carrying capacity K, and is negative above K.rKNR(N)

Growth is positive below the carrying capacity K, zero at K, and negative above K.

Population trajectories

Solutions of the logistic equationPositive populations below K rise toward K. Populations above K fall toward K. A population equal to K stays constant.N₀ < KN₀ > KKtN(t)

K is a stable equilibrium: every positive trajectory shown approaches the carrying capacity.

Two D3 graphs showing how the logistic model changes sign at carrying capacity K and how positive solutions approach K.

Solving the logistic equation

For a nonconstant solution with N≠0N\ne0N=0 and N≠KN\ne KN=K, separate the variables:

KN(K−N) dN=r dt.\frac{K}{N(K-N)}\,dN=r\,dt. N(K−N)K​dN=rdt.

The partial-fraction decomposition is

KN(K−N)=1N+1K−N.\frac{K}{N(K-N)} =\frac1N+\frac1{K-N}. N(K−N)K​=N1​+K−N1​.

The second antiderivative has a minus sign because the derivative of K−NK-NK−N is −1-1−1:

∫(1N+1K−N)dN=ln⁡∣N∣−ln⁡∣K−N∣=ln⁡∣NK−N∣.\begin{aligned} \int\left(\frac1N+\frac1{K-N}\right)dN &=\ln|N|-\ln|K-N|\\ &=\ln\left|\frac{N}{K-N}\right|. \end{aligned} ∫(N1​+K−N1​)dN​=ln∣N∣−ln∣K−N∣=ln​K−NN​​.​

Therefore,

ln⁡∣NK−N∣=rt+C,\ln\left|\frac{N}{K-N}\right|=rt+C, ln​K−NN​​=rt+C,

and the sign can be absorbed into one arbitrary constant C1C_1C1​:

NK−N=C1ert.\frac{N}{K-N}=C_1e^{rt}. K−NN​=C1​ert.

If N(0)=N0N(0)=N_0N(0)=N0​ and N0≠KN_0\ne KN0​=K, then

C1=N0K−N0.C_1=\frac{N_0}{K-N_0}. C1​=K−N0​N0​​.

Solving for NNN gives

N(t)=KN0N0+(K−N0)e−rt.\boxed{ N(t)=\frac{KN_0}{N_0+(K-N_0)e^{-rt}} }. N(t)=N0​+(K−N0​)e−rtKN0​​​.

This formula describes the nonconstant positive solutions. The values excluded during separation must be checked separately. They give the two equilibrium solutions

N(t)≡0andN(t)≡K.N(t)\equiv0 \qquad\text{and}\qquad N(t)\equiv K. N(t)≡0andN(t)≡K.

For r>0r>0r>0:

  • if 0<N0<K0<N_0<K0<N0​<K, then N(t)N(t)N(t) increases toward KKK;
  • if N0=KN_0=KN0​=K, then N(t)N(t)N(t) stays at KKK;
  • if N0>KN_0>KN0​>K, then N(t)N(t)N(t) decreases toward KKK.

Thus KKK is both the carrying capacity and the stable positive equilibrium.

A falling object with linear drag

Take downward as the positive direction. Let:

  • v(t)v(t)v(t) be the downward velocity;
  • m>0m>0m>0 be the object's mass;
  • g>0g>0g>0 be gravitational acceleration;
  • γ>0\gamma>0γ>0 be the linear drag coefficient.

Gravity contributes +mg+mg+mg, while drag opposes the velocity and contributes −γv-\gamma v−γv. Newton's second law gives

mv′(t)=mg−γv(t),mv'(t)=mg-\gamma v(t), mv′(t)=mg−γv(t),

or

v′(t)=g−γmv(t).\boxed{v'(t)=g-\frac{\gamma}{m}v(t)}. v′(t)=g−mγ​v(t)​.

This is a first-order linear equation. Put it in standard form:

v′+γmv=g.v'+\frac{\gamma}{m}v=g. v′+mγ​v=g.

Its integrating factor is

μ(t)=eγt/m.\mu(t)=e^{\gamma t/m}. μ(t)=eγt/m.

Multiplying the equation by μ\muμ creates a product derivative:

ddt(eγt/mv(t))=geγt/m.\frac{d}{dt}\left(e^{\gamma t/m}v(t)\right) =ge^{\gamma t/m}. dtd​(eγt/mv(t))=geγt/m.

Integrating gives

eγt/mv(t)=mgγeγt/m+C.e^{\gamma t/m}v(t) =\frac{mg}{\gamma}e^{\gamma t/m}+C. eγt/mv(t)=γmg​eγt/m+C.

If v(0)=v0v(0)=v_0v(0)=v0​, then C=v0−mg/γC=v_0-mg/\gammaC=v0​−mg/γ, so

v(t)=mgγ+(v0−mgγ)e−γt/m.\boxed{ v(t)=\frac{mg}{\gamma} +\left(v_0-\frac{mg}{\gamma}\right)e^{-\gamma t/m} }. v(t)=γmg​+(v0​−γmg​)e−γt/m​.

Velocity under linear drag

Falling-object velocity curvesVelocities above terminal velocity decrease toward it. Zero and negative initial velocities increase toward the same terminal value.mg/γterminal velocityv₀ above terminal velocitynegative v₀tv(t)
The exponential term decays, so every curve approaches the terminal velocity mg/γ.

Since m>0m>0m>0 and γ>0\gamma>0γ>0,

e−γt/m⟶0as t→∞.e^{-\gamma t/m}\longrightarrow0 \qquad\text{as }t\to\infty. e−γt/m⟶0as t→∞.

Therefore,

lim⁡t→∞v(t)=mgγ.\boxed{ \lim_{t\to\infty}v(t)=\frac{mg}{\gamma} }. t→∞lim​v(t)=γmg​​.

The value mg/γmg/\gammamg/γ is the terminal velocity. It is also a stable equilibrium:

  • an initial velocity above mg/γmg/\gammamg/γ decreases toward it;
  • an initial velocity below mg/γmg/\gammamg/γ, including a negative one, increases toward it;
  • an initial velocity equal to mg/γmg/\gammamg/γ remains constant.

Source: https://notes.ohevan.com/notes/ode/03-logistic-and-falling-models

© 2026 Evan Luo. All rights reserved.

Back to Ordinary Differential Equations

0 reads

  • stay up to date

  • about me

  • coffee

© 2026 Evan Luo. All rights reserved.