Tropical cyclone activities at ECMWF

ECMWF has a wide range of users of forecasts of tropical cyclones (TCs). Several member states have territories that are frequently hit by TCs, and Météo-France is the responsible Regional Specialized Meteorological Centre (RSMC) for the southern Indian Ocean. ECMWF forecasts of TCs are also routinely used by WMO member states.  Additionally, ECMWF recently opened the operational charts to the public.  Many commercial customers have activities related to TCs. 

Traditionally, forecasters and users have relied on global models including ECMWF to predict the position (or “track”) of the centre of the TC out to 5 days.  Nowadays, users are increasingly demanding a wider range of products related to TCs, and NWP centres are working to meet their needs. These needs include seasonal forecasts of TC activity, sub-seasonal forecasts of the potential for TC genesis (formation), accurate medium-range forecasts of the intensity and structure of each TC, and downstream influences of TCs on extratropical weather (including over Europe).  The interest in forecasting impacts of TCs has also increased.  At ECMWF, the Copernicus Emergency Management Service (CEMS) run global flood forecasts. Given that an important fraction of flood events in the tropics are related to TCs, accurate forecasts of the magnitude, structure, and duration of rainfall from TCs are necessary.  Several external institutes also use ECMWF forecasts to model storm surges induced by TCs; storm surges have historically been the largest cause of death from TCs.

Within this perspective of ever-increasing user needs, this report has been developed to document progress and challenges in the ECMWF forecasting system of special relevance to TCs. These included observations (Section 2), tracking of TCs (Section 3), verification (Section 4), predictions (Section 5), data assimilation (Section 6), modelling (Section 7), forecast products (Section 8), and impact forecasting applications (Section 9). Of note, especially in Sections 6 and 7, are new results from a series of special experiments conducted over a common 37-day period in August-September 2020 that explored potential future avenues in data assimilation and modelling.  Section 10.1 summarizes key progress and challenges in each of the above areas, and Section 10.2 provides a larger-scale vision for future improvement. As this report was spawned by a 1-year visit by Prof. Sharan Majumdar from University of Miami, and due to the unusually high activity in the Atlantic basin during 2020, a special attention is paid to the performances in the Atlantic. The breadth of this report has resulted in a large amount of material. To improve its readability, the report is written in a way that each section can be treated as a stand-alone article on a specific topic, summarizing progress, challenges, and future directions.