Journal of Climatology & Weather Forecasting

ISSN - 2332-2594


Verification of Ensembled Forecasting

Varsha Duhoon, Rashmi Bhardwaj

Ensemble based prediction systems over the years have proved their worthiness in providing better forecast guidance. The systems have evolved to regional scale from global scale since past twenty years. The objective of reducing error in forecast has led to ensemble approach applied at convective scale resolution. NCMRWF Regional Ensemble Prediction System (NEPS-R) is based on the regional version of Met Office Global and Regional Ensemble Prediction System (MOGREPS) with 12 members (1 control+11 perturbed). This regional ensemble prediction system having 4 km horizontal resolution and 80 vertical levels up to a height of 38.5 km is implemented in “Mihir HPCS”. NEPS-R uses boundary conditions and initial conditions generated from NEPS-G. Uncertainties of model are managed using Random Parameters (RP) scheme. NEPS-R is aimed at providing 3-day probabilistic forecasts using 12 members. Results of Bias and RMSE of the Global, Regional (1+11) have been compared and analyzed to understand the climate condition in India for the month of August 2019. NEPS-R predicted wind at U 850 and rainfall and compared with its global counterpart NEPS-G in predicting severe weather phenomena.