One of the challenges of generating regional analysis and forecasting is achieving
a proper specification of boundary conditions. In particular, for a regional data
assimilation system, it is important to study the impact of different ways of
generating boundary conditions. Given the way the model data is relaxed to boundary conditions, it can lead to a degradation in the quality of the analysis
and therefore the forecasts. In this work, a possible treatment is proposed to
overcome these difficulties by examining the sensitivity in the analyses and
forecasts of a regional data assimilation and forecasting ensemble system, the
Local Ensemble Transform Kalman Filter – Weather Research and Forecasting
Model (LETKF-WRF), including GEFS global model’s information as a boundary
condition and through the spectral nudging technique (SN). Numerical experiments
were carried out in a period of 2 months, evaluating the impact of different factors
of the SN technique in the analysis and forecasts generated, using a multi-scheme
ensemble of 20 members, composed of combinations between cumulus and planetary
boundary layer parameterizations. The results show that the implementation of the
SN together with the data assimilation system has a positive impact, improving
the performance of the circulation and thermodynamic variables. Furthermore, it
is emphasized that the implementation of this technique is feasible and has great
potential to improve regional analysis and forecasts, which should continue to be
studied in depth with new experiments.