Time 00:00Outside 0 CNest air 0 CVent flow 0 L/sCO2 0 ppmFPS 0

Drag to orbit, wheel or pinch to zoom. Cut face colour = temperature (blue cool, red hot). Dots in the cavity are air: red leaving at the top (warm nest), blue moving down (cool nest). Insects are drawn 8x larger than life.

Termite Mound Wall as a Thermal Buffer: Day-Night Conduction, Stack-Effect Ventilation and Nest CO2 Balance

A savanna mound is cut open. Heat enters through the mud wall with a delay, the nest air follows, and the buoyant flow through the vents reverses between day and night, flushing the CO2 the colony makes. Change the wall, colony and vents and see what stays steady.

Weather

Mound and colony

Readout

Temperatures over the last 48 h (C)

Temperature through the wall: now, and daily min/max (C)

x axis: depth from the outer surface in cm. The envelope narrows with depth as exp(-x/delta); delta = sqrt(2 alpha / omega) is the penetration depth of the daily wave.

Vent flow (L/s, positive = out at top) and CO2 (ppm / 100)

Experiment: how thick must the wall be?

Each row is a separate 6-day run (last day measured) at the current weather without sun heating, so the outer forcing is a pure harmonic. Predicted = A exp(-d/delta), a semi-infinite estimate for the inner wall surface; it fails for thin walls.

Independent numerical checks

Controls

Weather sets the mean and daily swing of outside air and the extra solar heating of the outer wall. Wall thickness, colony size, vent area, chimney height and wind change the mound.
Time runs 30 min to 6 h per real second. Reset starts again at midnight with an equilibrated mound.
View buttons jump to the whole cut-away, the wall section or the outside.

The science

Wall conduction. Heat flow through the mud wall follows ∂T/∂t = α ∂²T/∂x², solved on 36 nodes with implicit Euler. For a harmonic surface temperature of angular frequency ω (one day) the wave decays as exp(-x/δ) with δ = √(2 α/ω) and lags by x/δ radians. With α = 5x10^-7 m²/s (typical dry mud) δ is about 12 cm, so a wall 30 cm thick cuts the daily swing to under 10 %.

Nest air. A lumped balance: C dT_in/dt = Q_met + hA (T_wall - T_in) - ρ c_p |Q| (T_in - T_out), with interior thermal mass C = 2.5x10⁵ J/K, film coefficient h = 3 W/m²K, wall area 12 m² and 20 microwatts per termite (all typical or illustrative values).

Stack effect. A temperature difference between nest and outside drives a buoyant flow Q = C_d A_v √(2 g H |ΔT| / T) (C_d = 0.6); a small wind-driven term 0.04 A_v u is added. The sign flips when the nest becomes warmer or cooler than the air outside: warm air leaves at the top at night, and the flow reverses by day, as in measurements on real mounds.

CO2 budget. V dC/dt = G + |Q| (C_out - C), with production G from the metabolic heat (respiratory quotient 0.85, 21 kJ per litre of oxygen) and C_out = 420 ppm. At steady flow C = C_out + G/|Q|.

Simplifications. One-dimensional wall, one lumped nest volume, no moisture, no soil below, no wind pressure field, no porosity of the wall (real mounds exchange gas through it), no shading by grass. The numbers are order-of-magnitude illustrations of the physics, not measurements of a species; real nests often hold 1 to 3 % CO2. The 3D insect models are enlarged 8x.