Sensitivity to initial conditions and horizontal resolution in a case study of intense precipitation over Argentina

DILLON, MARIA EUGENIA; KALNAY, EUGENIA; GARCIA SKABAR, YANINA; RUIZ, JUAN; COLLINI, ESTELA

Florianopolis

Simposio; Seventh International WMO Symposium on Data Assimilation; 2017

One of the big challenges in numerical weather prediction is to reduce the uncertainty in the initial conditions. Another one, is to represent all the dynamical and physical processes that take place on the atmosphere with an adequate resolution. At the National Meteorological Service (SMN) of Argentina many efforts have been carried out to progress in both directions. Experiments on a regional data assimilation system are being developed using the Local Ensemble Transform Kalman Filter (LETKF) coupled with the Weather Research and Forecasting Modeling System (WRF) with a relatively low resolution, and at the same time a high resolution deterministic WRF is run everyday in a semi operational schedule. In this work, the intense precipitation event of 6th December 2012 has been selected to evaluate the impact in the numerical forecasts of using different initial conditions or different resolution. For the LETKF-WRF experiments, horizontal resolution of 40 km and 40 ensemble members are used, assimilating the observations available in the PREPBUFR files from the Global Data Assimilation System (GDAS-NCEP). First, we delve into the sensitivity of this system to distinct configurations: a multi scheme structure combining cumulus and planetary boundary layer parameterizations for the ensemble members; the inclusion of the temperature and humidity vertical profiles retrieved from the Atmospheric Infrared Sounder (AIRS) in the assimilation cycle; the use of theoretical perturbed lateral boundary conditions. In addition, deterministic experiments are carried out using the Global Forecasting System (GFS) as the initial condition of the WRF, with resolutions of 40, 27 and 4 km.A verification of accumulated precipitation forecasts is presented through measures as the Equitable Threat Score (ETS), Probability of detection (POD), False alarm ratio (FAR), Bias Score (BIASS) and Probability Distribution Function (PDF). The results indicate that both the use of a data assimilation system to improve initial conditions and the increase in the resolution, cooperate to obtain better forecasts. Then, it would be necessary to progress in the development and evaluation of a higher resolution data assimilation system. In order to benefit from such a system, the inclusion of radar data would be a must.