DATE2018-05-18 03:59:54
IDABSTRACT20180518035954-0200
CONTACTtbarrera@meteo.cat
PRESENTATIONORAL
INVITED0
IDSESSION3
TITLE1-KM CLIMATE PROJECTIONS FOR CATALONIA BASED ON STATISTICAL DOWNSCALING FROM CMIP5 MODELS
AUTHORSA Barrera-Escoda (1), V Altava-Ortiz (1,2), J Amaro (1), J Cunillera (1), M Prohom (1), A Saironui (1)
AFFILIATIONS
  1. Meteorological Service of Catalonia, Barcelona, Catalonia (Spain)
  2. Forest Fire Prevention Service, Santa Perpètua de la Mogoda, Barcelona, Catalonia (Spain)
ABSTRACTTo assess future climate change impacts in Catalonia, a climatic diverse territory located at the North-western Mediterranean basin, high-resolution (1 km) temperature and precipitation climate projections for the period 2006-2100 have been developed. Future projections have been obtained using a statistical downscaling technique applied to three CMIP5 models (MPI-ESM, GFDL-ESM2G and CanESM2), and forced by three RCP emission scenarios (RCP2.6, RCP4.5 and RCP8.5). These models have shown a good skill on reproducing past synoptic variability in the North Atlantic area and Western Europe. The downscaling method is based on the meteorological analogues concept which takes advantage of a daily high-resolution (1 km) temperature and precipitation database developed from a high-densely weather station network for the period 1971-2015, which has been subjected to a primary quality control (removal of outliers). This method shows an outstanding improvement of the inter-annual cycle and variability range of precipitation and temperature in comparison of those obtained from dynamical downscaling simulations in the study area. The technique also tries to solve the issue on projecting extreme temperature values, usually a weakness in the statistical downscaling methods based on the analogy concept. The obtained projected changes in temperature and based climate indices are almost linearly dependent with increasing CO2 concentration. In detail, temperature projections show a clear increase of its annual-mean values (2-6ºC respect to 1971-2000, during this century). However, this increase is not homogeneously distributed within the year. The highest warming is found in autumn and spring, supporting the idea of an extension of summer-like conditions. Magnitude of projected change principally depends on the distance from the sea and altitude. On the other hand, extreme maximum and minimum temperature indices changes are heavily triggered with the CO2 concentration at the end of the century. An increase up to 60 and 30 days on average during this century are expected in the number of warm days (TX>30ºC) and tropical nights (TN>20ºC), respectively. As regards precipitation projection, it is pointed out a general decrease on annual-mean values but with a great range of variability. Seasonally, it is found a significant decrease during summer and a lesser one for spring, while winter and autumn do not present any staidly significant trend. It is also obtained a significant increase of high-precipitation events (>50 mm/day) and a decrease of low-precipitation days (<5 mm/day).
PAGE72
STATE1