Atmospheric convection as an unstable predator-prey process with memory

Abstract Heuristic models and observational analyses of atmospheric convection often assume that convective activity, for example rain rate, approaches some given value any large-scale (“macrostate”) environmental conditions such as static stability humidity. We present novel convection-resolving simulations in which the activity evolves a fixed equilibrium mean state (“macrostate”). In this case is unstable, diverging quasi-exponentially away from either to extreme or zero rate. Thus almost rate can coexist with an profile: model also depends on history. then two-variable, predator-prey motivated by behavior, wherein small-scale (“microstate”) variability produced by, but promotes precipitation, while macrostate properties CAPE promote, are consumed precipitation. model, influenced much its own history (via persistent microstate variability) current environment. This reproduces simulated instability found above could account several lag relationships observed convection, including afternoon maximum over land well-known “quasi-equilibrium” balance at synoptic time scales between forcing response key variables. These results point strong role memory suggests basic strategies observing, modeling parameterizing processes should pay closer attention smaller than grid box.