New Improvements to the ECMWF Forecasting Package for Summer 2018

(credit: European Centre for Medium-range Weather Forecasting)

DISCUSSION: With ever-increasing processing power being made available by the world’s top scientific research supercomputers, forecasters and researchers alike are keen to stay on the cutting edge of forecasting accuracy and reliability. The European Centre for Medium-range Weather Forecasting (ECMWF) has recently announced that several new upgrades and improvements are being introduced in the latest release of the ECMWF Integrated Forecasting Package (IFS) cycle 45r1. The IFS uses a sophisticated four-dimensional data assimilation scheme (e.g., latitude-longitude-altitude-time) and is fed information from a combination of observational data and model outputs which leads to generate high-resolution forecasts. Forecasts within the IFS are further separated into two classes: a high resolution forecast and an ensemble-based forecast.

New meteorological content is being introduced in IFS cycle 45r1 that will potentially enhance forecasting skill and quality. One of the biggest improvements in this cycle is the introduction of a three-dimensional coupled ocean-atmosphere scheme with data obtained from the Nucleus of European Modeling of the Ocean (NEMO) dataset version 3.4. Because the ocean and atmosphere work consistently in tandem, coupling of the ocean-atmosphere interface is important when considering accurate simulations of future conditions. The existing NEMO-IFS scheme has also been upgraded to allow for a full ocean-atmosphere coupling in the tropics, with partial coupling in the extratropics.

Another significant improvement is with regards to the bathymetry model, which has been upgraded to use the National Oceanic and Atmospheric Administration’s ETOPO1 (1 arc-minute) locked topography-bathymetry dataset. This dataset is a significant improvement from the predecessor dataset in that many biases in estimated bathymetrical depth have been corrected for and many erroneous measurements have been addressed. The importance of this is that the wave model within the IFS can tap into this improved data and forecast accuracy of wave heights can be greatly improved as a result. The figures at the top of this article shows the adjustments to the bathymetry with ETOPO1 data compared to the predecessor dataset in both the high-resolution and ensemble wave models.

So what do these improvements mean for us? Recalling the spirit of a coupled ocean-atmosphere interface, many improvements to the upper air forecasts are expected. Understanding more about our upper-air dynamics will provide more clues on predictability of. Near-surface temperature and precipitation biases also receive an improvement on the predecessor cycle, especially in the tropical regions and over Europe. On the tropical cyclone front, intensity error has been decreased by as much as 10% over the first 5 days of a forecast and up to a 20% reduction in error beyond day 5. This is an important topic that is stressed upon global forecasters for hurricane intensity changes, especially since rapid intensification processes in tropical cyclones continue to be a challenge for forecasters and researchers alike.

While it is still currently in the open testing phase, these new upgrades are expected to be released in just a few days (June 5th, 2018). For a complete description on the new improvements, additions, and preliminary findings with the new IFS cycle besides those mentioned in this article, check out the ECMWF documentation here.

Image credit: European Centre for Medium-range Weather Forecasting

To learn more about the latest meteorological research, click here!

© 2018 Meteorologist Brian Matilla