Table 1 Summary of numerical experiments.

From: Diurnal self-aggregation

Experiment
Name
Diurnal surface temperature
Amplitude, Ta [K]
Horizontal resolution
dx [km]
Domain size
L [km]
Days with
3D output
A5a 5 1 960 1–6
A2a 2 1 960 1, 4–6, 8
A5b 5 1 480 1–8
A2b 2 1 480 1–4, 8*, 9*
A5c 5 0.5 240 1–3
A5d 5 0.2 240 1–3
A3.5 3.5 1 480 3–5
A5sea 5 1 480 1–4
A2sea 2 1 480 1, 8
A5vent05 5 1 480 4
A5vent001 5 1 480 4
A5p2K 5 1 480 4, 8
A5ph3.6h 5 1.5 384 3–5
A5ph6h 5 1.5 384 3–5
A0 0 1 256 1–3
A5constrad 5 1.5 384 3–5
A0constrad 0 1.5 384 3–5
  1. The main four experiments are listed in the first four rows (A5a, A2a, A5b and A2b). The experiments labelled by a star (*) are equivalent to A2b but constitute an additional, longer-duration run (A2long). All the above experiments were carried out at Ts = 298 K, except A5p2K, which was carried out at Ts = 300 K. A5sea and A2sea were carried out, by assuming a water surface at the lower boundary. In all other experiments, surface evaporation was set to 70% of the value computed for a water surface. A5vent05 and A5vent001 denote experiments, where the ventilation coefficients52 were reduced to 0.5 and 0.01 of their default values. A5ph3.6h and A5ph6h constitute experiments, where the diurnal cycle of surface temperature was phase-shifted by 3.6 and 6 h. A5constrad and A0constrad are simulations, where the incoming solar radiation was set to a constant equalling its average value.