Downscaling climate change over French Guiana: from CMIP6 projections to sector-based indices

Over the past couple of decades, thanks to the sustained development of Global Climate Models (GCMs) combined with dedicated downscaling strategies such as regional climate modelling or statistical downscaling, climate projections and associated services are now increasingly available across many regions of Europe including France. However, whereas this holds for continental France, the national territory includes numerous overseas territories, some of them also EU outermost regions, where this information is still only partially available, if at all.

Here we present the results of statistical downscaling of Coupled Model Intercomparison Project phase 6 (CMIP6) GCMs for the small territory of French Guiana in equatorial South America. Compared with island territories, the larger size of French Guiana makes the direct use of GCMs possible, although statistical downscaling with long daily surface observations of temperature, humidity, precipitation and surface wind remains relevant for bias correction, mapping (the territory is covered by few grid points), and computing tailored climate indices for the agriculture, water resource, energy, or public health sectors.

Obvious advantages are cost-effectiveness and the processing of GCM ensembles that provide more reliable uncertainty estimates made difficult with the more expensive dynamical methods. However, systematic GCM biases such as a displaced Intertropical Convergence Zone may challenge model uncertainty assessment and rather call for a storyline approach supported by selected, skilled GCMs, which all project warming and drying trends over French Guiana. Among the various expected societal impacts, heat stress resulting from the combined humid climate and widespread warming strongly increases the risk of hyperthermia, and the energy demand for air conditioning as a likely local adaptation strategy.