Site-only notes on what the paper delivered.
What they got right (main result)
Geurts et al. (2026) give a clear G6-style comparison of hydrological and tropical-circulation responses to sulfur-based stratospheric aerosol injection versus three less longwave-absorptive solid particles — alumina, calcite, and diamond — in the fully coupled SOCOLv4 model under a transient SSP5-8.5 to SSP2-4.5 temperature target. Within that idealized framework, lower absorptivity reduces perturbations to the atmospheric radiation balance, lowers apparent hydrological sensitivity (precipitation change per degree of surface cooling), and limits lower-stratospheric warming and associated Hadley Cell and ITCZ shifts relative to sulfate. Regional precipitation and P–E patterns for the least absorptive materials sit closer to SSP2-4.5 in several tropical and extratropical metrics, and Northern Hemisphere summer monsoon drying is generally milder for less absorptive particles, while Southern Hemisphere monsoon responses stay limited. The authors also document the process reasons: tropical loading concentrates Fa and latent-heat reductions over the oceans, and absorptive LSW steepens stratospheric equator-to-pole gradients that feed dynamical side effects. They close with an explicit uncertainty fence — solid-particle simulations carry larger uncertainties than sulfur SAI — and they state the hydrological and dynamical benefit under the caveat that potential stratospheric interactions were neglected.
What they didn’t quite nail
Solid particles are injected as monodisperse aerosols at ideal scattering sizes, with sub-grid plume microphysics neglected, so agglomeration and cooling-efficiency losses inside real plumes are outside the experiment. Heterogeneous surface chemistry on solid particles and interactions with sulfuric acid are omitted, so the study does not constrain how coupled chemistry would reshape ozone, LSW, or the hydrological ranking. Aerosol–cloud interactions are absent; the paper calls for further study rather than quantifying whether clouds change material benefits. The design is tropical injection only, so claims about less-absorptive materials under off-equatorial or other strategies remain open, as do finer local and seasonal monsoon diagnoses for regional outliers.
Reader caveat (not a paper flaw): the low-absorptivity hydrological advantage is a constraint-call result under idealized SOCOLv4 assumptions — it is not proof the benefit survives once plume microphysics and heterogeneous chemistry are constrained, not a finding that clouds already offset the ranking, and not evidence that lower-absorptivity materials are ozone-safer or GLOMAP-ready. Any UKESM/GLOMAP path is outside this host.