Presets
Playback
Parameters
Camera
Layers
Quality
Model check
The science
What is modelled
A circular garden pond (radius 3.4 m, depth 1.3 m, about 47 m³) with a school of koi, lily pads, algae and a day-night cycle. Two models run together. The water chemistry is a classic well-mixed oxygen budget (Streeter-Phelps / Odum type): dissolved oxygen, algae and dissolved nitrogen change with the weather, the feeding and the fish. The koi are Reynolds boids whose tail-beat frequency follows the swimming Strouhal law. When oxygen falls below 4 mg/L the fish rise toward the surface and slow down.
Equations
- Oxygen saturation of fresh water at 1 atm (Benson and Krause, 1984):
ln Cs = −139.344 + 1.5757×105/T − 6.6423×107/T² + 1.2438×1010/T³ − 8.6219×1011/T⁴, T in kelvin (9.09 mg/L at 20 °C). - Oxygen:
dDO/dt = ka(Cs − DO) + 1.4(μ − ra)A − Rfish − SOD/h − 0.3 F/V, with algae biomass A, aeration rate ka (fountain, waterfall), koi respirationRfish = 3.6 · 2.2(T−20)/10 mfish/Vg O2/m³/day (about 150 mg O2 per kg per hour at 20 °C), sediment demand 0.25 g/m²/day (θ = 1.065) and the oxygen needed to digest feed F. - Algae:
dA/dt = (μ − ra − 0.05) A, growthμ = μmax 1.07T−20 N/(0.1+N) · I/(60+I)with μmax = 1.5 per day. The depth-averaged light I falls with the lily-pad cover c (90 % blocked under pads) and with the algae themselves (Beer-Lambert, kd = 0.5 + 0.12 A). - Nitrogen: feed adds 0.03 g N per gram, 0.06 g/m³/day comes from rain, top-up water and leaves; algae take up 0.08 g N per g; 0.03 per day leaves with water exchange.
- Light: a clear day,
I = 650 sin(π(h − 6)/12)W/m² between 06:00 and 18:00. The step is a 5-minute fourth-order Runge-Kutta; the oxygen terms are integrated alongside the state so the budget can be audited. - Swimming: cruise speed 0.8 body lengths per second at 20 °C with Q10 = 1.5; tail beat
f = St U / Awith St = 0.3 (the range 0.2-0.4 measured for swimming fish) and tail peak-to-peak amplitude A = 0.12 m (0.2 of the 0.6 m body).
Checks run in the page
- Saturation formula against the standard fresh-water table at five temperatures.
- Pure aeration against the exact exponential relaxation.
- Koi respiration against aeration: steady state Cs − D/ka.
- Oxygen budget closure: the integrated terms add up to the change of dissolved oxygen.
- Daily light dose against 2 Imax D / π.
- School containment, spacing and speed limit over 120 s.
Honest simplifications
- The pond is one well-mixed box: no stratification, no temperature profile and no wind-driven reaeration. The ka slider stands in for all surface exchange.
- All biological rate constants are textbook-range values (not fitted to one pond), so the numbers show how the budget behaves, not what your pond will do. Real ponds also have bacteria, nitrification, plants, pH and ammonia toxicity, none of which are modelled.
- The water temperature is a slider, not a heat budget; the day length is compressed so that one day lasts 20 to 240 s.
- The koi are not fed by the nitrogen model directly; pellets are thrown by hand and add 10 g worth of nitrogen and oxygen demand. Fish growth, disease and the real stress limits of koi (they differ by individual) are not modelled; 4 mg/L is an indicative stress level and 2.5 mg/L an indicative danger level.
- Fish movement is 2-D steering with a depth layer; real schooling involves vision and the lateral line, and koi are not strongly schooling fish. The surface ripples and the algae colour of the water are visual.
What to try
- Choose "Hot summer, overfed": watch oxygen fall each night until the koi crowd the surface, then pick "Same pond with a fountain".
- Lower the temperature to 6 °C: saturation rises to 12.5 mg/L and the koi barely move (about 0.5 body lengths/s).
- Raise the lily-pad cover to 70 %: less light reaches the algae and the day-night swing flattens.
- Throw pellets and watch the school accelerate toward them.
Loading model...
Loading Blender model...