Hot weather impacts on urban indoor air temperature assessed through citizen science observations in the Netherlands

Due to climate change and urbanization, the world's population is increasingly exposed to extreme heat, posing a threat to public health. Despite people spending ~90% of their time indoors, heat effects in buildings have been studied far less than outdoor heat island effects. This study aims to observe, understand and model the behaviour of indoor air temperatures (T_in) during summer heat. As a proof of concept, we present and analyse up to 27 years of individual T_in timeseries of seven citizen weather stations (CWS) across the Netherlands. First, we find that typically T_in increases slower, but also cools down slower than T_out with a lag difference of ~130 minutes in the diurnal cycle. We demonstrate that nocturnal indoor human thermal comfort (HTC) can be worse than outdoor HTC even for days after a heatwave.

Second, to model T_in behaviour, we simulate six-hour changes in T_in behaviour with a physics-based statistical model by Vant-Hull et al. (2018) that has an outdoor conduction, indoor conduction and solar transfer component. Preliminary results of this computationally-fast model for each of the seven houses are promising, showing on average a R-squared of 0.74 and a root mean squared error of 0.13 K. Third, in the next research steps, we are also interested in how T_in may evolve due to climate change. We will study this by converting the T_in measurements to 2050 and 2085 values based on the Royal Netherlands Meteorological Institute 2014 climate scenarios (or 2023 if available).

Finally, we will scale up our proof-of-concept analyses to 100 indoor CWS placed in Amsterdam. The participating households receive a CWS for three years to measure their indoor climate – temperature, relative humidity, CO₂ concentrations – in the bedroom and living room. Based on our insights, we will make recommendations for climate-sensitive urban design to reduce indoor heat stress.