Comparing anthropogenic heat input and heat accumulation in the subsurface of Osaka, Japan

In metropolitan areas, shallow groundwater temperatures are affected by anthropogenic heat sources. The resulting thermal conditions in the subsurface are highly site-specific, and spatial and temporal trends have only been revealed for a few cities. In this study, the anthropogenic heat input is quantified for 15 locations in Osaka, Japan using an analytical, one-dimensional conductive heat transport model. Mean anthropogenic fluxes into the subsurface are determined annually between 2003 and 2011. The model depicts fluxes from buildings and from different land cover types separately. The main objective is to compare the predicted annual mean heat input to heat storage increase, and to identify site-specific factors relevant for the thermal evolution of the underground at each well location. Our results indicate that mean fluxes from asphalt covered areas (0.28 ± 0.07 W/m2) and from buildings (0.32 ± 0.18 W/m2) are significantly higher than fluxes from unpaved (0.06 ± 0.06 W/m2) and grass-covered (−0.04 ± 0.06 W/m2) areas. Furthermore, the temporal variation of mean fluxes from buildings is stable over the studied time period, while annual mean fluxes from asphalt, grass and unpaved areas vary as much as 0.8 MJ/m2. Still, the uncertainty associated with the combined annual heat input of all heat sources is slightly higher than the changes between the years. Overall, the predicted cumulative heat input (2003 to 2011) at the wells ranges from 4 MJ/m2 to 60 MJ/m2. Comparing these results to heat storage increase, additional local heat fluxes, such as from construction work or a sewage treatment plant, have to be considered for about 1/3 of the wells. In addition, it becomes apparent that a significant percentage of determined anthropogenic heat input is not stored in the urban aquifer and heat input is predicted to be considerably higher than heat storage increase. [Display omitted] •Heat flux into the underground is quantified for 4 heat sources and 15 wells.•Heat flux in Osaka is considerably lower than previously determined flux in Europe.•Sealed surfaces and buildings contribute equally to urban subsurface warming.•Temporal analysis reveals little variation in annual heat input.•For undisturbed wells heat accumulation correlates to heat input, but is lower.