Report from CLARIS WP3.1: Climate Change Downscaling Partners: CNRS, CONICET, UBA, ¿IMPE?, ¿USP?, INGV, UCLM, UCH, MPI External partners: SENAMHI (Peru), Univ. Reading (UK)

Objectives To make-up a SA-EU expert group on dynamical and statistical downscaling techniques. To set up a research strategy on regional-scale climate change in SA. To assess the performance of RCMs to reproduce current regional climate in SA. To develop scenarios for the late 21 st century using different downscaling methods. To analyse the changes in mean climate, weather regimes and extremes over different South American regions. To compare the suitability of the employed downscaling methods for regional climate change impact studies in SA. To develop standard RCM simulated data sets.

Month 3: Setting-up of a common research strategy by the joint research group on downscaling methods in southern South America T1: To assure a unified approach for all experiments in terms of model domain and resolution, time periods, model forcing and diagnostics studied through our SA-EU expert group on dynamical and statistical downscaling techniques. To define experiments T2 and T3 T2: To explore the capability of the RCMs, forced by reanalysis data, to simulate some case-studies (synoptic time-scale) of daily temperature and/or precipitation extreme events already described in the litterature (link with WP3.2). T3: To perform and analyze present-day RCMs simulations driven both by reanalysis data and by the available ESMs, focusing on their ability to simulate extremes of temperature and precipitation (link with NCT2 and WP3.2). Month 12: 2 nd workshop to discuss i) the results of the case-studies (T2), ii) the results of the seasonal time-slice “perfect boundary” experiments for RCMs in order to assess their ability to reproduce the observed climate, and iii) the definition of experiments T3 and T5 (RCMs simulations driven by ESMs, including statistical and statistical-dynamical downscaling). Evaluation of the strategy and reorientation, if necessary, of its priorities. Milestones and expected results

Common research strategy on downscaling methods in SSA

Dynamical downscaling: Models: MM5 (CONICET), MM5 (UCH), RAMS (SENAMHI), REMO (MPI), PROMES (UCLM), LMDZ (CNRS) Domain: 0º – 50ºS, 35ºW – 85ºW Horizontal resolution: 50 – 60 km

1 st year: Simulations: All of RCMs nested in ERA40 (3 months run) Periods: To be defined (3 events in La Plata Basin + 1 event in Chile) Goal: To test if main mesoscale processes responsible for climate extreme events (heat waves, floods, etc.) in SSA are “captured” by RCMs Timetable: ERA40 boundary data set must available for all RCM groups before December 2 nd year: Simulations: All of RCMs nested in ERA40 (3 years run) Goal: To assess the performance of RCMs to reproduce interannual variability in SA. Simulations: Some RCMs (¿PROMES?, ¿INPE-RCM?) nested in any ESM (30-years control+scenario)

Statistical downscaling: Goals: - To test how local climates in La Plata Basin are represented by statistical downscaling techniques - To compare statistical and dynamical downscaling results Predictor data set: ERA40 atmospheric fields Predictants data set: Daily Tm, Tx, Tn, Pr (¿other surface vars?) observed in La Plata Basin stations 1 st year: Boulanger & Penalba (CNRS-UBA): Neuronal networks 2 nd year: Penalba (UBA), Coelho (UR): Regression - correlation Vera & Vargas (CONICET-UBA): Weather regimes (PCA)

D3.1 (~month 4): Report of the 1 st workshop giving a general outlook over ongoing work on South American regional climate modeling and downscaling Deliverable Published papers related with : a)Regional atmospheric models in SA b)Statistical downscaling in SA Climate downscaling work accomplished in SA What more...? D3.2 (~month 12): Report of the 2 nd workshop analysing the preliminary results about the capability of the RCMs (driven by reanalysis) to simulate extreme events and the present-day climatology of southern South America.