Imagine you're measuring how fast a river flows by counting how many boats pass a bridge per hour. The river slows down, and fewer boats pass — but then you realize the boats just got smaller, so more fit in each stretch of water, and the actual water volume hasn't changed at all. You were measuring the container, not the flow. That's what happened to the stability-iris hypothesis. The committed claim: the widely cited mechanism linking surface warming to reduced tropical anvil cloud amount — via increasing static stability aloft — is an artifact of pressure-coordinate formulations, not a real dynamical constraint on cloud formation. The apparent stability increase with warming comes from decreasing air density at upper levels, which changes how things look in pressure coordinates without changing how convection actually builds and sustains anvil clouds. This is not a minor technical quibble — it dissolves a core physical argument that has shaped how the climate community thinks about cloud feedbacks. McGraw, Gasparini, Mol, and Voigt demonstrate this through two channels. First, an analytical decomposition that isolates the density-reduction contribution to the stability trend, showing it dominates. Second, they verify with an ensemble of RCEMIP (Radiative-Convective Equilibrium Model Intercomparison Project) simulations that while the stability increase is indeed robust across models, neither anvil cloud amount, convectively driven water convergence into anvil levels, nor the efficiency of that convergence in producing cloudiness shows a systematic response. The stability metric goes up; the clouds shrug. The ladder here is unusual — this paper is not trying to beat a benchmark but to dismantle a theoretical expectation. The stability-iris hypothesis, most associated with Bony et al. and subsequent work building on the FAT (Fixed Anvil Temperature) framework, predicted that as tropical sea surface temperatures rise, anvil clouds would thin out because they'd rise into more stable environments. Multiple GCMs have shown ambiguous or contradictory anvil responses to warming, but the community lacked a clean explanation for why the simple physical argument didn't hold. This paper provides that explanation: the "stability" being invoked is substantially a coordinate artifact. Architecturally, this is analytical thermodynamics plus multi-model ensemble verification — not a machine-learning or simulation-heavy approach. The core move is a decomposition of the static stability tendency in pressure coordinates into a genuine lapse-rate component and a density-reduction component, then showing the latter dominates. The RCEMIP ensemble provides the simulation evidence that models agree on the stability increase but diverge on cloud response, which is exactly what you'd expect if the stability signal is physically inert for cloud formation. Integrity is strong for this class of paper. RCEMIP is a community intercomparison project — these aren't the authors' own simulations cherry-picked for agreement. The analytical derivation is checkable by anyone with a thermodynamics textbook. The result is negative (dispelling an expectation rather than confirming one), which carries lower publication-bias risk. The main caveat: this addresses whether basic-physics arguments predict systematic anvil reduction — it does not rule out that anvil clouds might still change with warming through other mechanisms. The obvious next experiment the authors did not run: testing whether alternative formulations of the stability-iris mechanism — ones that don't rely on pressure-coordinate static stability — can rescue any version of the hypothesis. The honest read is (c): this is the clean demolition paper, and the reconstruction paper is the next publication. They've cleared the ground; someone (possibly them) will now need to ask what actually does control anvil cloud amount trends with warming.
