This page is still under development.

The case studies are being prepared. In the meantime, this interactive tool shows how I turn a technical model into something you can test in the browser.

INTERACTIVE / ILLUSTRATIVE DISPERSION MODEL

Methane Field Lab

Place sources and monitors, steer the wind, then collect synthetic observations from a model you can inspect.

COMPUTING IN YOUR BROWSER

Source flux

Wind + dispersion

Multiplies both lateral σy and vertical σz spread. It is an illustrative dispersion control, not a molecular diffusion coefficient.

Snapshot observation

SourcesContinuous monitors
Move S1

125 m east · 210 m north

Drag on the field. Keyboard: focus the field and use the arrow keys.

Total simulated flux24 kg/h
Monitors intercepting plume1 / 2
Fixed heatmap scale0–12 ppm
600 × 400 m simulated field■ source · ● monitor · • snapshot sample

1 sources emit 24 kilograms of methane per hour in total. The displayed wind is from 270.0 degrees at 3.0 meters per second. The diffusion scale is 1.00 times the neutral baseline. 1 of 2 continuous monitors intercept the modeled plume.

Drag any source or monitor to move it. Keyboard: select an object, focus the field, and use the arrow keys.

CONTINUOUS NETWORK

What the monitors observe

Concentration is measured; emission rate is inferred through the assumed plume model.

MonitorCH₄ enhancementPlume statusModel-implied combined flux
M12.87 ppmIntercepted23.7 kg/h
M20.000 ppmOutside modeled plumeNot estimable
LIVE CONTINUOUS MONITORS

Record the monitor network

PAUSED0 / 120 OBSERVATIONS8 READINGS / SECOND1 POINT = 5 SIMULATED SECONDS
M1 waiting M2 waiting
Press Play to begin collecting simultaneous observations from every continuous monitor.

Every update advances a seeded, mean-reverting random walk in wind direction and speed around your selected conditions, then re-evaluates the steady-state plume and adds seeded Gaussian monitor noise. The chart keeps the latest 120 observations and drops the oldest point as each new one arrives. It is a teaching stream, not a puff model, atmospheric forecast, or validated monitor specification.

SYNTHETIC OBSERVATION

Take a measurement snapshot

Drone transect: illustrative 0.15 ppm detection limit and 0.08 ppm baseline Gaussian noise. These are teaching assumptions, not vendor specifications.

Choose a technology and take a snapshot to generate synthetic data.

Teaching model only. It assumes continuous point releases, flat unobstructed terrain, a neutral-stability dispersion baseline with a user-controlled spread multiplier, constant wind, and no chemistry, buoyancy, deposition, or building effects. A real inversion requires calibrated instruments, defensible meteorology, uncertainty propagation, and validation against controlled releases.

WHAT THIS MODEL DOES

A small inverse problem, made visible.

The field uses a steady-state Gaussian plume to translate source flux, wind, and distance into a methane concentration enhancement. Move a monitor through the plume and it observes concentration, not emission rate. The displayed rate is inferred by running that concentration back through the same assumed transport model.

C(x, y, z) ∝ Q / (u σy σz)

That distinction matters: a sensor can report what passed its location, while source quantification additionally depends on meteorology, geometry, transport assumptions, and uncertainty. Snapshot technologies add synthetic noise and detection limits so a visitor can see how the latent plume becomes an imperfect dataset.

MODEL CONTRACT

Useful for intuition.
Not a field protocol.

This first release deliberately uses the simplest continuous point-source model: fixed wind, a neutral atmospheric-stability baseline, flat terrain, and no buildings, chemistry, deposition, or plume rise. The diffusion control scales the baseline lateral and vertical plume widths; it is not a molecular diffusion coefficient. The sensor profiles are illustrative teaching assumptions, not vendor specifications.

NOAA's Air Resources Laboratory documents the Gaussian plume model for continuous point-source dispersion. EPA's OTM-33A materials show how Gaussian point-source concepts can support methane quantification, with much more operational detail than this demonstration.