Using the Firefly optimization method to weight an ensemble of rainfall forecasts from the Brazilian developments on the Regional Atmospheric Modeling System (BRAMS)

摘要

In this paper we consider an optimization problem applying the metaheuristicFirefly algorithm (FY) to weight an ensemble of rainfall forecasts from dailyprecipitation simulations with the Brazilian developments on the RegionalAtmospheric Modeling System (BRAMS) over South America during January 2006.The method is addressed as a parameter estimation problem to weight theensemble of precipitation forecasts carried out using different options ofthe convective parameterization scheme. Ensemble simulations were performedusing different choices of closures, representing different formulations ofdynamic control (the modulation of convection by the environment) in a deepconvection scheme. The optimization problem is solved as an inverse problemof parameter estimation. The application and validation of the methodology iscarried out using daily precipitation fields, defined over South America andobtained by merging remote sensing estimations with rain gauge observations.The quadratic difference between the model and observed data was used as theobjective function to determine the best combination of the ensemble membersto reproduce the observations. To reduce the model rainfall biases, the setof weights determined by the algorithm is used to weight members of anensemble of model simulations in order to compute a new precipitation fieldthat represents the observed precipitation as closely as possible. Thevalidation of the methodology is carried out using classical statisticalscores. The algorithm has produced the best combination of the weights,resulting in a new precipitation field closest to the observations.