Performance of RegCM4 convective permitting version in precipitation extremes over southeastern South America: preliminary results Rosmeri P. da Rocha *, M. Llopart, M. L. Bertolli, Fernandez J., Chou S., Coppola E., Doyle M., Feijoo M., Gutierrez J.M., Lavin-Gullon A., Solman S. * Universidade de São Paulo
CORDEX-FLAGSHIP PILOT STUDY IN SOUTHEASTERN SOUTH AMERICA (SESA) The CORDEX Flagship Pilot Studies (FPS) are targeted experimental setups to better addressing key scientific topics MOTIVATIONS Extreme precipitation events over SESA are • becoming more frequent and more intense. These events have large socio-economic and hydrologic impacts. There is a need for better understanding and • modeling these precipitation extremes. There are limited ESD studies in the region. • There is a need for developing RCM and ESD • coordinated actions. Courtesy of M. L. Bertolli
CORDEX-FLAGSHIP PILOT STUDY IN SOUTHEASTERN SOUTH AMERICA (SESA) MAIN OBJECTIVES: to study multi-scale processes and interactions • (convection, local, regional and remote processes, including the co-behaviour of processes) that result in these extreme precipitation events; and to develop actionable climate information • from multiple sources (statistical and dynamical downscaling products) based on co-production with the impact and user community. Courtesy of M. L. Bertolli
CORDEX-FLAGSHIP PILOT STUDY IN SOUTHEASTERN SOUTH AMERICA (SESA) Models ARGENTINA Department of Atmospheric and Ocean Sciences – University of Buenos Aires Partners ARGENTINA WRF CIMA-National Council of National Council of Scientific and Technical Research (CONICET) ARGENTINA Argentine Association of Regional Consortia for Agricultural Experimentation (AACREA) BRAZIL RegCM4 University of São Paulo (USP) BRAZIL São Paulo State University (UNESP) BRAZIL Center for Weather Forecasting and Climate Studies (CPTEC) BRAZIL ETA National Institute for Space Research BRAZIL Brazilian Agricultural Research Corporation (EMBRAPA) CZECH REPUBLIC Charles University in Prague SPAIN WRF CSIC / University of Cantabria ITALY Abdus Salam International Centre for Theoretical Physics (ICTP) URUGUAY Department of Atmospheric Sciences, Courtesy of M. L. Bertolli University of the Republic
Objective and strategy As part of CORDEX-FPS the aims of this study is to understand the ability of RegCM4 convective permitting version to simulate precipitation extremes over SESA; Strategy: three case studies within one of the warm seasons with higher number of extreme events: from November 2009 to March 2010; Total number of lightnings from 1998-2011 observed SESA by LIS. ( Anselmo, 2015) (a) Figura 13: N´ umero total de raios ( flashes ) observados pelo LIS em cada regi˜ ao de 0,25 ¶ × 0,25 ¶ de latitude e longitude entre 1998–2011, ilustrando a matriz FL lis .
Test Protocols: 1. Three extreme events: Case1: 18-02-2010 00:00 UTC - event peaks 20/21-02-2010 stations max: 127 mm/day Case2: 17-01-2010 00:00 UTC - event peaks 19-01-2010 station max: 152 mm/day Case3: 20-11-2009 00:00 UTC - event peaks 22-11-2009 station max: 115 mm/day 2. Two simulation types Weather mode (WM) – simulation starts ~12 hours before initial phase of each one of three extreme events; Climate mode (CM) – continuos simulation (seasonal) starting at 01-10-2009 ending at 31-03-2010.
Test Protocols: 3. Two domains : ~ 24 (CSAM-20i) and 4 (SESA-4i) km of grid spacing; 4. Initial and boundary conditions: CSAM-20i experiments are nested in ERA-Interim reanalysis; SESA-4i experiments are nested in CSAMi-20i; 5 . Models : RegCM4, WRF and ETA. 6. RegCM4 non-hydrostatic core RegCM4 physic options: CSAM-20i CSAM-20i : uses Tiedke and Kain-Fritsch convective schemes, respectively, over the land and ocean; SUBEX for large-scale precipitation SESA-4i : WSM5 microphysics scheme (convective permitting) SESA-4i CLM4.5: to solve surface processes in both simulations
Climate mode (CM) simulations (CSAM-20i x SESA-4i) : Climatology for Nov/2009-March/2010 TRMM Simulations capture the CSAM-20i observed maximum of rainfall in southeast SA; More intense precipitation is simulated by CSAM-20i CPC Impact of convective SESA-4i permitting simulation: Decreases in the intensitity of the seasonal rainfall à values closer to the observation.
Climate mode climatology: November/2009-March/2010 CSAM-20i x SESA-4i - zoom CSAM-20i Monthly mean precipitation CSAM-20i SESA-4i TRMM SESA-4i CPC Month-to-month variability is similar in both simulations; Higher agreement with observations in convective permitting simulation: spatial pattern and intensity;
CM simulations of the precipitation diurnal cycle (mm/day) Long Simulations - Diurnal Cycle Climatology for Nov/2009 to March/2010 18 16 14 12 mm/day Stations 10 CSAM-20i 8 SESA-4i 6 Location of the TRMM 4 stations from 2 INMet/Brazil 0 3 6 9 12 15 18 21 0 Time (UTC) Stations: precipition peaks at 12 and 15 UTC (9 and 12 LT) à it is correctely reproduced by both simulations; Simulations overestimate the intensity of rainfall during the day; During the nigth (21, 0, 3 UTC) à intensity biases are small Most of the times SESA-4i (convection permiting) is closer of local observations than CSAM-20i;
Cases study: Climate mode x weather mode Parametrization (CSAM-20i) x coventive permitting (SESA-4i) simulations
Case1: evolution of observed daily precipitation from 19 to 21 February 19 20 21 February CMORPH CPC CMORPH hourly rainfall Very intense Precipitation Pcp continues pcp (above 120 initiates intense in mm/day) CMORPH from from 18 to between 19-20 20 to 21 19 February February stations max: 127 mm/day
Case1: simulations x observation Simulations are able to reproduce the CMORPH general aspects (location and intensity) of the extreme event February 20 CSAM-20i SESA-4i CM: location and area of the more intense rainfall are better reproduced by convective permintting (SESA-4i) CM simulation; WM: more intense rainfall is slightly southwestward shifted compared to CMORPH in CSAM-20i; WM SESA-4i simulates better the location, but underestimates the area with intense rainfall.
Case3: evolution of observed daily precipitation from 21 to 22 November 21 22 November CMORPH Precipiation intensifies on November 22 CPC: Between 100-120 mm/day CMORPH: above 120 mm/day CPC
Case3: simulations x observation Precipitation above 120 mm/day is simulated by both CSAM-20i and SESA-4i CSAM-20i SESA-4i CM : location of more intense precipiation is closer to the CMORPH in convective permiting (SESA-4i) simulation; WM: simulations displace northward the maximum rainfall compared to CMORPH
Case3: precipitation diurnal cycle Diurnal Cycle_Case3 80 70 60 mm/day 50 Stations 40 30 CSAM-20i 20 SESA-4i 10 0 3 6 9 12 15 18 21 0 Time (UTC) WM (Weather Mode simulations) Simulated and observed precipiation peaks at 15 UTC (12 LT); Simulations overestimate the intensity during most of the times; During the day (9-18 UTC) à overestimation is greater in convective permitting (SESA-4i) than in CSAM-20i.
Preliminary conclusions and next For one of warm seasons with higher number of extreme events (Nov/2009- March/2010): the two simulations captured the general aspects of the seasonal precipitation; Convective permitting simulation presents greater agreement with observations than parameterized one, including the diurnal cycle of precipitation in (CM). Initial results with convective permitting RegCM4 version are encouraging to study extreme events over Southeast South America. Next: - to quantify the improvement of the simulations using the convective permitting version (statistical index, PDFs, diurnal cycle, etc.); - to compare RegCM4 simulations with other participant models (WRF and ETA); - to understand the environment conditions favoring these extreme events.
Thanks! Obrigada!
Recommend
More recommend
