INVESTIGADORES
GALLIGANI Victoria Sol
congresos y reuniones científicas
Título:
Forecast Evaluation of a Deep Convection Case During Relampago Assimilating Conventional and Satellite Observations with the WRF-GSI-LETKF System
Autor/es:
CORRALES, PAOLA BELEN; JUAN RUIZ; GALLIGANI, VICTORIA
Lugar:
Bonn
Reunión:
Simposio; WCRP-WWRP Symposium on Data Assimilation and Reanalysis; 2021
Resumen:
We present forecast evaluation results from the first attempt to use the WRF-GSI-LETKF system in Argentina. The impact of assimilating high-resolution surface networks and satellite observations on a forecast from the analyses is evaluated. This study also represents the first steps to assimilate satellite radiances in a regional context in South America. We focus on a case study corresponding to a huge mesoscale convective system (MCS) developed over central and north-eastern Argentina during November, 22th, 2018. Analyses with 10-km horizontal grid spacing were produced assimilating observations every hour from 11/20 18Z to 11/23 12Z. We used a 60-members ensemble which at the first assimilation cycle is initialized from the deterministic GFS run adding random perturbations with climatological covariance. A multiphysics approach is also used to represent model errors, using different physics configurations. Four assimilation experiments were conducted using different sets of observations: CONV with conventional observations from prepBUFR, AUT that uses CONV observations plus automatic station networks, SATWND add satellite-derived winds and RAD includes satellite radiances from AMSU, HIRS, MHS, ATMS, AIRS and IASI. We run ensemble forecasts initialized from the analyses every 6 hours between 11/22 00Z and 11/22 06Z to evaluate the impact of the different observing networks on forecasts. Comparing the forecasts with the observed precipitation (IMERG Final  Run), the representation of precipitation is shown to improve for forecasts initialized from AUT, SATWND and RAD. Moreover, the impact continues for many hours after the initialisation of the forecasts. Complete results and conclusions from these experiments will be presented at the symposium.