Equatorial waves and the skill of NCEP and ECMWF numerical weather prediction systems
In a new study to be published in Monthly Weather Review, CIRES and NOAA researchers at the Physical Sciences Division and collaborators examined models from the European Centre for Medium-Range Weather Forecasts (ECMWF) and NOAA National Center for Environmental Prediction (NCEP). They found that, despite progress in observations, forecast models, and computing, critical deficiencies in both systems’ tropical forecast skill remained when comparing rainfall forecasts to those at higher latitudes. The ECMWF tends to perform better than the NCEP around the globe and especially over the tropics. Their study also suggests that a major source of these differences stems from limitations in model physics related to clouds and rainfall.
This study shows that large-scale tropical weather disturbances in the ECMWF are more strongly coupled to rainfall than in the NCEP system. Because a substantial amount of tropical rainfall is organized by such disturbances, ECMWF rainfall predictions tend to be more skillful than NCEP’s. By analyzing forecast skill by lead time, the authors show that there is a positive relationship between the skill of northern latitudes forecasts beyond the fourth day and the skill of the tropical forecasts at the first day.
This study highlights the current gap in skill between ECMWF and NCEP by focusing on tropical rainfall forecasts. In addition, this work implies that model development aimed at improving tropical rainfall variability is important not only to improve regional forecasts, but can also potentially benefit longer-term weather forecasts outside of the tropics in both hemispheres.