Global urban climate analysis on multi-decadal km-scale climate simulations

Current research on future urban climate involves several steps from global climate simulation planning (e.g. CMIP), through downscaling and down to impact modelling. This process is often time-consuming and can introduce physical and technical inconsistencies due to the variety of tools and assumptions involved. With the advent of coupled km-scale multi-decadal modelling, pioneered in the nextGEMS project and operationalized by Destination Earth, the workflow can now be accelerated and simplified by directly performing urban climate analysis on the global simulations.

We present what is, to our knowledge, the first analysis of urban climate on global km-scale climate simulations. We exploit the fully coupled IFS-FESOM and IFS-NEMO simulations produced on the nextGEMS project. These span for several years across horizontal resolutions of 28, 9 and 4.4 km, and include a 30-year historical simulation (1990-2020) at 9 km, as well as a 30-uear future scenario (scenario 2020-2050) following the SSP3-7.0  scenario.

We present a robust automated methodology that, based on surface static information, allows to gather and filter simulation information over all kinds of urban areas and their rural references worldwide. We then show the spatiotemporal features of such urban and rural model representations consistently for hundreds of cities worldwide. We further analyze the sensitivity of the nextGEMS present-day multi-annual simulations to spatial resolution, with a particular focus on the thermal aspects of urban climate and its contrast to the rural counterpart in each season. By exploiting km-scale surface temperature remote sensing products by LSA-SAF, we can confirm that the models are able to capture key spatial and temporal features of the typical surface urban heat island diurnal cycle, and that performance improves with increasing spatial resolution.

Finally, we make use of the available multi-decadal simulations by IFS-FESOM to explore the evolution of urban climates for both past and future scenario and group urban areas according to their expected change. We further investigate the possible amplifiers, such as orography or urban density, in an attempt to explain the different trends for cities worldwide found in the simulations.