Climate-induced Risk Assessment of Library Collections within Dora I WWII Bunker in Trondheim, Norway

Short- and long-term variability in indoor microclimate conditions within conservation spaces (museums, galleries, archives, and libraries) can exacerbate the risk of deterioration of cultural materials. Active microclimate control systems are often installed to stabilize thermo-hygrometric conditions, conversely, massive buildings with thick walls are less affected by outdoor conditions due to their high thermal inertia. Historic reinforced concrete structures, like bunkers and fortifications, are often windowless, further providing stable microclimate and protecting sensitive materials from photodegradation. World War II (WWII) bunkers often labeled as “dark” or “painful heritage”, were initially perceived only as symbols of war and occupation, leading to neglect and material deterioration over time. However, the impossibility of their demolition gradually encouraged interest towards their reuse, raising awareness of their social, historical, and economic potential. This contribution provides a comprehensive microclimate analysis to support management strategies in a unique case study: the NTNU (Norwegian University of Science and Technology) library for cellulose-based materials, hosted within “Dora I” WWII bunker in Trondheim, Norway (63.43° N 10.40° E). The archive spans 3700 m² and contains around 5200 m³ of collections, including ancient volumes, journals, newspapers, and pictures. “Dora I”, a massive reinforced concrete German submarine and terrestrial fortification with 3.5 m-thick walls, covers a total area of approximately 16,000 m². 10 thermo-hygrometers compliant with European standards were installed, ensuring representativeness of indoor conditions across two floors. Time series of air temperature (T) and relative humidity (RH) are significantly longer (7 years, since 2018) than most studies in the literature, and are continuous and complete with minimal number of missing values. Statistical approaches from climatology are applied to analyse T-RH data, decomposing time series into short-term (daily/noise) and long-term (seasonal) variability (extracted by sinusoidal fits) to provide insights into indoor climate dynamics. A comprehensive conservation risk assessment, based on dose-response functions, evaluated biological and chemical threats to archival materials using both raw and filtered (i.e., “clean”) microclimate data. This approach allows to examine how anthropogenic factors (e.g., access and archival management) in the noisy signal may exacerbate climate-induced conservation risks. The findings demonstrate the exceptional features of this massive building, where peak summer temperatures occur indoors 2 to 2.5 months later than outdoors, depending on sensor location. The study estimates no biological risks for cellulose materials from humidity-dependent insects or mould growth, with temperature-dependent insect degradation that remains a threat, particularly from June to December (although it has decreased over time). Chemical degradation risks, confirmed by three indicators, remains significant from July to October. This is evident when considering raw microclimate data, which reveals slight but still significant variations in risky days compared to "clean" data, suggesting a potential influence of human activities related to archival management. In conclusion, the findings underscore the benefits of massive structures in preserving vulnerable materials and a useful methodological approach in combining raw and filtered microclimate data to assess conservation risks. Analysing noise signal may inform conservators about the impact of their management practices, offering a useful framework for similar archival contexts worldwide.