Future climate projections for PAs: Climate Downscaling

26-27 September 2016, Pisa, Italy

The general strategy for future climate projections in ECOPOTENTIAL Protected Areas was discussed and decided during the Workshop on Future Projections held in Pisa, Italy, on September 26–27 2016. More than 20 researchers attended the meeting, representing ten protected areas within the project: Hardangervidda, Camargue, Curonian Lagoon, Gran Paradiso, Kruger, La Palma, Lake Orhid, Negev, Sierra Nevada and Wadden Sea. Researchers illustrated their needs in terms of future climate data needed as inputs for modelling and understanding changes in ecosystems and ecosystem services in the PAs and discussed the best strategy to adopt together with expert partners in climate modelling (CNR).

The strategy that will be adopted includes the use of specific downscaling methods for climate variables, in particular precipitation and air temperature, to be applied to coarser resolution data, such as those produced by regional climate models, in order to generate high resolution climatic inputs for the eco-hydrological models applied in the different PAs.

For both air temperature and precipitation, the downscaling tools that are being employed to generate the fine scale climatic fields take into account the orography. This is a necessary requirement for a downscaling method if it has to be applied in complex areas, such as mountainous or coastal areas. For precipitation, a stochastic downscaling method built on the RainFARM technique previously developed at CNR improved with the implementation of the orographic correction is employed. For temperature, the downscaling method takes into account bias correction procedures and an orographic correction based on the temperature lapse rate. After testing their performance, the methods have been applied to the outputs of Regional Climate Models from the EURO-CORDEX experiment at ~11 km spatial resolution over one selected PA, the Gran Paradiso National Park. An example of temperature downscaling over the Gran Paradiso National Park with initial data at 11 km resolution and final data at 1 km resolution is shown in the following figure.

An important outcome of the Pisa Workshop was the identification of the model data to be downscaled. In agreement with the project partners, the Regional Climate Model datasets included in the EURO-CORDEX archive was selected. It was also decided to focus first on the data having the highest spatial and temporal resolution (~11 km and 3-hr respectively) a requirement which is fulfilled by the RCA4 Regional Climate Model driven by five different Global Climate Models from the CMIP5 ensemble. For future projections, the decision was to employ the standard RCP4.5 and RCP8.5 scenarios from the latest IPCC Assessment Report.