Second campaign of microclimate monitoring in the carcer tullianum: temporal and spatial correlation and gradients evidenced by multivariate analysis

BACKGROUND: This paper discusses results obtained in the second monitoring campaign of the Carcer Tullianum, a particular hypogeum environment located in the historical centre of Rome (Italy). In the first paper we stressed the need to apply chemometric tools to this kind of studies in order to obtain full and significant information; really information on sampling design, sensors (type, number, position) and instrument validation seems to be not easy to find in literature for researches dealing with monitoring of indoor environments. Also in this case three main parameters (temperature, humidity, illuminance) were monitored in the complex construction by an inexpensive self-assembled system along some horizontal and vertical vectors together with some measurements of oxygen, carbon dioxide and barometric pressure. With respect to the first campaign, we used a higher number of sensors to cover a new excavated zone; for the same reason, as well as to take into account the presence of visitors, a different experimental design was adopted. RESULTS: Different data treatments were applied to data coming from all the used sensors. A good view of the microclimate was obtained that also resulted coherent with the different position of the three rooms constituting the monitored site (Carcer, Tullianum, Convent). Classical time plots resulted useful to evidence the correlation of the main monitored parameters (T, RH% and illuminance) with macroclimate, as well as their delay in following macroclimate. Box-Whisker and Gain-Loss graphs evidenced at the best the microclimate differences between the three rooms; an almost hypogean microclimate was evidenced for the lower room (Tullianum) where humidity values range between 90 and 100% while lower values, but anyway higher than the external, and spread more widely were measured passing to Convent and Carcer with minimum values around 50% for the last. A scarce or very scarce correlation with macroclimate was evidenced for all the three main measured parameters. Lighting results mainly dependent on artificial light and only in few cases, but unfortunately in the most precious zone, illuminance exceeds values suggested by Normative. CONCLUSIONS: Box-Whisker and Gain-Loss graphs allowed us to have the best view of the microclimate for all the monitored rooms. The influence of lighting by lamps on the other monitored parameters resulted overlapped and clearly topped the solar one. The worst situation was found in the Carcer, where the presence of the main chandelier worsens the state of the frescoed walls, already subjected to wide changes in temperature and humidity. Also the lighthouse located above the Convent provokes lighting exceeding values suggested by Normative while, as expected, LEDs resulted as suitable source of light from a conservation point of view. Susanne Heidi Plattner, Patrizia Fortini and Maria Pia Sammartino contributed equally to this work