Watch multiple walkers diffuse in real time. The MSD log-log plot reveals the diffusion law. Switch between lattice, Gaussian, and Lévy flight walks.
The core measure of a random walk is the Mean Squared Displacement (MSD). It quantifies how far the walkers have wandered, on average, after a certain time.
$$\langle r^2(t) \rangle = 2 D t$$Here, $\langle r^2(t) \rangle$ is the mean squared displacement at time $t$, $D$ is the diffusion coefficient (which depends on step size and rate), and the factor 2 is for two dimensions. This linear relationship defines normal diffusion.
To identify different types of motion, we often use a power-law form and analyze it on a log-log plot.
$$\langle r^2(t) \rangle \propto t^{\alpha}$$The exponent $\alpha$ is the anomalous diffusion exponent . On the simulator's log-log plot, the slope of the MSD curve is $\alpha$. If $\alpha = 1$, it's normal diffusion. If $\alpha \lt 1$, it's subdiffusion (trapped, crowded). If $\alpha \gt 1$, it's superdiffusion (directed or Lévy flight).
Pollution & Contaminant Spread: Modeling how a pollutant disperses in groundwater or air is a classic random walk problem. Engineers use these simulations to predict contamination plumes and design remediation strategies, often needing to account for anomalous diffusion in porous soils.
Financial Market Modeling: The random walk hypothesis suggests stock price movements are unpredictable, similar to a particle's Brownian motion. "Lévy flight" models, which you can simulate here, are used to better capture the rare, large price jumps (crashes or rallies) seen in real markets.
Animal Foraging Patterns: Many animals, from albatrosses to deer, do not search for food in simple Brownian paths. Biologists use Lévy flight models (switch the simulator to this mode) to describe patterns with many short moves and occasional very long jumps, which is an optimal search strategy in sparse environments.
Polymer Physics & Materials Science: The conformation of a polymer chain in a solution can be modeled as a random walk. Understanding whether its motion is normal or subdiffusive (slowed by obstacles) is key for designing gels, plastics, and pharmaceutical delivery systems.
First, it's easy to think that "the diffusion coefficient D is a constant determined solely by the type of particle," but this is a misconception. For example, even the same molecule will diffuse at completely different speeds in water versus in oil. In this simulator, try setting the "Walker Type" to "Continuous," keeping the "Step Size" fixed, and moving the "Diffusion Coefficient" slider. You'll see the "liveliness" of the particle's motion change, right? This means the diffusion coefficient depends strongly on the "environment (e.g., viscosity)" and "temperature." In practice, the first step is correctly setting the value of D used in your simulation based on experimental or literature values.
Next, do not judge the MSD graph based on a single run. Especially when the number of walkers is low, the MSD plot can be quite scattered. If you run a "Lattice" walk with just 1 walker, you should see that the MSD does not form a perfect straight line. This is statistical fluctuation. To obtain reliable results, you need to run the simulation with a sufficient number of walkers (e.g., 100 or more) and take the average.
Finally, be careful when setting parameters for Lévy Flight. If you set the power exponent α to a value close to 2.0 (e.g., 2.1), it will appear almost indistinguishable from a normal random walk. To clearly see the characteristics of superdiffusion, set α between 1.5 and 2.0. However, if you set α too low (e.g., 1.1), extremely long jumps can occur, potentially causing the particle to immediately leap out of the simulation area. When applying this to real phenomena, a process of carefully estimating α from observational data is essential.
Simulate colloidal gold nanoparticles (50 nm diameter) in water at 25°C. Set N=1000 particles, step size 0.5 μm, drift 0 μm/s (pure diffusion), Gaussian mode. After 500 time steps (Δt=1 ms each), MSD reaches approximately 2.5 μm². From Einstein relation D = MSD/(4t), calculated diffusion coefficient D ≈ 1.25×10⁻¹² m²/s, matching Stokes–Einstein prediction for this particle size and medium viscosity.