The abstract from your article:
Cloud cover and relative humidity in the tropics are strongly influenced by organized atmospheric convection, which occurs across a range of spatial and temporal scales. One mode of organization that is found in idealized numerical modeling simulations is self-aggregation, a spontaneous transition from randomly distributed convection to organized convection despite homogeneous boundary conditions. We explore the influence of domain geometry on the mechanisms, growth rates, and length scales of self-aggregation of tropical convection. We simulate radiative-convective equilibrium with the System for Atmospheric Modeling (SAM), in a non-rotating, highly-elongated 3D channel domain of length > 104 km, with interactive radiation and surface fluxes and fixed sea-surface temperature varying from 280 K to 310 K. Convection self-aggregates into multiple moist and dry bands across this full range of temperatures. As convection aggregates, we find a decrease in upper-tropospheric cloud fraction, but an increase in lower-tropospheric cloud fraction; this sensitivity of clouds to aggregation agrees with observations in the upper troposphere, but not in the lower troposphere. An advantage of the channel geometry is that a separation distance between convectively active regions can be defined; we present a theory for this distance based on boundary layer remoistening. We find that surface fluxes and radiative heating act as positive feedbacks, favoring self-aggregation, but advection of moist static energy acts as a negative feedback, opposing self-aggregation, for nearly all temperatures and times. Early in the process of self-aggregation, surface fluxes are a positive feedback at all temperatures, shortwave radiation is a strong positive feedback at low surface temperatures but weakens at higher temperatures, and longwave radiation is a negative feedback at low temperatures but becomes a positive feedback for temperatures greater than 295–300 K. Clouds contribute strongly to the radiative feedbacks, especially at low temperatures.
****************************************************************
The entire abstract is talking about self-aggregating convection, not temperatures. And it's all addressed to something seen in "IDEALIZED NUMERICAL MODELS".
Can you not read?