Open AccessQPF performance of the updated SIMM forecasting system using reforecasts
Author(s) -
Mariani Stefano,
Casaioli Marco,
Lanciani Alexandre,
Flavoni Simona,
Accadia Christophe
Publication year - 2015
Publication title -
meteorological applications
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.672
H-Index - 59
eISSN - 1469-8080
pISSN - 1350-4827
DOI - 10.1002/met.1453
Subject(s) - environmental science , meteorology , quantitative precipitation forecast , forecast verification , advection , grid , climatology , precipitation , computer science , forecast skill , geology , geography , physics , geodesy , thermodynamics
ABSTRACT In late 2009, the meteorological component of the Institute for Environmental Protection and Research ( ISPRA )'s Hydro‐Meteo‐Marine Forecasting System ( Sistema Idro‐Meteo‐Mare – SIMM ), namely the hydrostatic Quadrics BOlogna Limited Area Model ( QBOLAM ), was replaced by an up‐to‐date version of the same model. This new version, which is simply referred to as BOLAM , deploys advanced parameterization schemes for cumulus convection, radiation, soil, and exchange of turbulent fluxes and uses the explicit advection of five hydrometeors. As a part of the SIMM verification programme, it was judged objectively how these changes affected the quality of quantitative precipitation forecasts ( QPFs ). This was done by intercomparing the current operational BOLAM against QBOLAM . Observational analyses, from April to September 2001, deployed earlier to evaluate the QBOLAM performance, were considered for this intercomparison. An ad hoc reforecast campaign was performed to generate the database of the corresponding retrospective QPFs for BOLAM . A multimethod verification approach was used to assess the spatial scales resolved by the intercompared model versions and the skill at forecasting both frequent, high‐base rate and rare, low‐base rate precipitation events. The effect on performance measures of differences in bias between the intercompared versions was also evaluated. The results indicate an improved performance of BOLAM compared with the older version not only when considering the native 0.1° model grid, where the intercomparison may be affected by differences in the resolved scales, but also when using a coarser 0.5° grid, where the two remapped QPF series show similar spatial scales. Fewer false alarms are produced by BOLAM during the verification time period.