Field Evaluation Of Attic–Indoor Temperature Dynamics Beneath A Lightweight Metal Roof In A Hot-Humid Tropical Climate, Owerri, Nigeria

Abstract

Lightweight metal roofs are widely used in hot-humid residential environments, yet long-term field evidence on the relationship between attic and occupied-space air temperatures beneath such roof configurations remains limited. This study evaluated the diurnal and seasonal attic–indoor temperature dynamics of a residential building in Owerri, Nigeria, comprising a black stone-coated metal roof, an attic with no purpose-provided ventilation, and a plaster-of-Paris ceiling. Attic and occupied-space air temperatures were synchronously monitored from 2 October 2024 to 22 June 2025 at six-hour intervals, yielding 1,052 paired observations. Descriptive statistics, temperature-amplitude analysis, Pearson correlation and ordinary least-squares regression were used to characterise the relationship between the two monitored environments, with Newey–West heteroskedasticity- and autocorrelation-consistent standard errors applied to regression inference. Mean attic air temperature was 31.15 °C compared with 30.57 °C in the occupied space, while temperature amplitudes were 7.40 °C and 5.85 °C, respectively, corresponding to an attic-to-indoor temperature-amplitude ratio of 0.79. Mean attic–indoor temperature difference decreased from 0.66 °C during the dry-season period to 0.40 °C during the wet-season onset. Attic and occupied-space temperatures were strongly positively correlated (r = 0.956, p < 0.001), and the regression model showed a high degree of shared linear variation (R² = 0.913). These findings provide field-based observational evidence of strong but temporally variable thermal coupling between the attic and occupied space for the investigated roof–attic–ceiling configuration. They further indicate the relevance of considering roof solar heat gain, attic ventilation and ceiling thermal resistance within an integrated building–environment approach to passive thermal design. The findings are configuration-specific and do not establish attic temperature as an independent cause of occupied-space thermal conditions.