Heat today, gone in a fortnight? Event-time evidence of attitudinal attenuation following heat wave exposure

Public concern about climate change often rises during heat waves and then fades, yet its temporal profile remains under-specified. We link monthly CIS barometers in Spain to province-day temperature records and operationalize exposure as the number of days since the most recent local threshold-heat day (TMAX > 34°C). Our exposure-linked repeated cross-section design estimates event-time effects by grouping interviews according to how recently the local threshold-heat day occurred, from the same day to short lags of one to two weeks, to medium lags of two to four weeks, and to longer lags extending over one to three months and beyond. The preferred no-burden CRE/Mundlak model, which retains non-accumulated heat controls but excludes cumulative burden measures, reveals no clear post-event decline in the first two weeks once covariates are introduced, but a modest negative association from roughly day 16 onward. This pattern weakens materially under stricter calendar absorption using year-month and exact-date fixed effects, which counsels caution against strong mechanistic claims. Exploratory subgroup models suggest an earlier and steeper decline among men than among women. Ideology and gender display strong expected gradients. The findings are consistent with a fading-affect interpretation but do not rule out seasonal or media-driven alternatives. The study documents a short-lived recency pattern in climate concern rather than a large attitudinal shift, and suggests that interview timing relative to recent extreme heat can matter for survey-based estimates of climate concern.