In 2019, ECMWF continued to provide high-quality weather predictions to its Member and Co-operating States and other users of its data and products across the globe. Severe weather events that were well predicted by ECMWF forecasts include freezing rain early in the year in Romania, two short heatwaves that brought record-breaking temperatures to many parts of Europe, and tropical cyclones that hit the Azores and Mozambique.

An upgrade of the Integrated Forecasting System (IFS) in June improved the skill of forecasts substantially across most variables and regions. Amongst many other improvements, IFS Cycle 46r1 introduced more continuous data assimilation to improve ECMWF’s estimate of the state of the Earth system at the start of forecasts. It also included new ocean wave physics and new output parameters in the extended range to provide better advance information on the probability of severe weather.

Among many improvements in forecast scores, ECMWF recorded its highest ever skill for Extreme Forecast Index wind predictions. ECMWF also widened access to its products by making Ocean5 reanalysis charts freely accessible, and more products were made available to Members of the World Meteorological Organization (WMO). Experimental products to predict cold spells in Europe were made available to registered users, and a new product generation package was rolled out.

Extreme Forecast Index (EFI) and Shift of Tails (SOT) forecasts
Extreme Forecast Index (EFI) and Shift of Tails (SOT) forecastsThe charts show the EFI (shading) and SOT (contours) for maximum temperature in forecasts from 00 UTC on 26 June for 26–28 June (top left); from 00 UTC on 23 July for 23–25 July (top right); from 00 UTC on 20 June for 26–28 June (bottom left); and from 00 UTC on 17 July for 23–25 July (bottom right). The white square in the top left panel shows the area of 44°N–50°N and 2°E–8°E referred to in the text.

Two European heatwaves

In 2019, national all-time maximum temperature records tumbled in Belgium, France, Germany, Luxembourg, the Netherlands and the UK. The records were broken during two relatively short but extreme episodes of heat that hit western Europe at the end of June and the end of July 2019. The geographical extent of the heatwaves was well captured a week in advance, as illustrated in the panels showing Extreme Forecast Index predictions 7–9 days ahead.

In the medium range, the ensemble forecast for the area shown as a white square in the EFI plots was more confident about the extreme heat for the June episode. This can be seen in the two plots below, where the blue symbols indicate the range of possible outcomes predicted at different times ranges ahead of the event.

However, between 9 and 11 days before the June event, most ensemble members actually predicted a cold anomaly. For the July episode, a warm anomaly was present in forecasts issued as early as two weeks in advance. Later on, the ensemble distribution continued to smoothly shift to more and more extreme temperatures, but with a large ensemble spread.

Ensemble temperature forecasts
Ensemble temperature forecastsThe charts show the evolution of forecasts for 3-day average 2-metre temperature in western-central Europe (the white square in the EFI plots) valid on 26–28 June (left) and 23-25 July (right). The blue box-and-whisker symbols show ensemble forecasts for different starting dates. The red dots indicate ECMWF’s deterministic high-resolution forecasts (HRES).

Response to tropical cyclone Idai

The tropical cyclone season of 2019 in the southern Indian Ocean was one of the most active on record, with 15 tropical storms. On 15 March, tropical cyclone Idai made landfall in Mozambique, causing around a thousand fatalities. This made Idai the deadliest cyclone in the southern Indian Ocean for more than 100 years.

ECMWF’s tropical cyclone track forecasts are available as free and open data. Our forecasts predicted the landfall location and extreme precipitation and winds with high confidence about 5 days ahead of landfall. At the same time range, flood forecasts based on ECMWF’s precipitation forecasts indicated a moderate risk of severe flooding, rising to a very high risk after landfall.

ECMWF worked with partners at the Universities of Reading and Bristol and the UK Government to support the humanitarian response to the disaster.

It did so by helping to provide scientific information on flood hazard and population exposure. Forecasts from the Copernicus Emergency Management Service’s Global Flood Awareness System (GloFAS), for which ECMWF is the computational centre, played a key role. They were used in combination with satellite imagery and flood extent maps to identify where, when and for how long flooding may occur, as well as where people may be impacted.

ECMWF’s track, position and intensity forecasts for Idai
ECMWF’s track, position and intensity forecasts for IdaiThe plots show forecasts starting at 00 UTC on 7 March 2019 (left) and 00 UTC on 10 March 2019 (right). The squares indicate position and intensity forecasts for 12 UTC on 14 March. The black line shows the observed track (‘Best Track’ estimate), while the observed position and intensity at 12 UTC on 14 March is indicated by the hourglass symbol.

New forecast outputs

The upgrade of ECMWF’s Integrated Forecasting System to IFS Cycle 46r1 in June 2019 introduced an Extreme Forecast Index (EFI) for water vapour flux as well as new EFI and Shift of Tails (SOT) products to highlight potential extremes in the extended range. The new extended-range EFI and SOT were for 2-metre temperature and total precipitation. An example is shown in the figure.

Also as part of 46r1, probabilities for 850 hPa temperature anomalies in terms of standard deviations from the model climate average, as well as additional probability thresholds for precipitation and near-surface (10 m) wind, were added to support the activities of World Meteorological Organization Members. Ocean fields, including sub-surface data such as the depth of the 20°C isotherm and the average salinity and potential temperature in the upper 300 m, were also made available.

Extended-range Extreme Forecast Index (EFI) & Shift of Tails (SOT)
Extended-range Extreme Forecast Index (EFI) & Shift of Tails (SOT)The chart shows the EFI (shading) and SOT (contours) for 2-metre temperature from 00 UTC on 3 June 2019 for the week from 10 to 17 June together with the locations where the observed 2-metre temperature in that week was below the 5th percentile or above the 95th percentile of the observed 20 year climatology. The forecast gave an indication of the anomalously cold conditions in the west and the unusually hot conditions in eastern Europe.

Forecast performance

On 11 June 2019, ECMWF implemented a substantial upgrade of its Integrated Forecasting System (IFS). IFS Cycle 46r1 included changes in the model and in the data assimilation procedure used to generate the initial conditions for forecasts. For example, the more continuous data assimilation introduced by 46r1 enabled the use of additional and more recent observations. The upgrade had a very positive impact on the skill of medium-range and extended-range ensemble forecasts (ENS) and medium-range high-resolution deterministic forecasts (HRES). Improvements could be seen across a range of parameters and atmospheric levels, including surface weather parameters.

Upper-air performance of the ensemble forecast (ENS) is monitored through the continuous ranked probability score (CRPS) for temperature at 850 hPa over the northern hemisphere extratropics. As the first chart shows, forecast skill during summer 2019 was higher than in previous summer seasons, and comparison with re-forecast skill based on the ERA5 reanalysis shows that this is partly due to the IFS upgrade.

The second score shown here monitors forecast skill for strong winds in terms of the Extreme Forecast Index. This score reached its highest ever value in 2019. There was also a slight increase in forecast skill for 2-metre temperature, and a gain in precipitation forecast skill. In addition, ocean wave parameters in the HRES were improved in Cycle 46r1 by up to 10% due to a major upgrade in the ocean wave model.

Skill of the ENS as measured by ECMWF’s primary headline score
Skill of the ENS as measured by ECMWF’s primary headline scoreEvolution of 850 hPa temperature ensemble forecast performance in the northern hemisphere extratropics, verified against the corresponding analysis. The chart shows 12-month and 3-month running average values of the forecast range at which the continuous ranked probability skill score (CPRSS) falls below 25%.
ENS headline score for strong winds
ENS headline score for strong windsEvolution of the skill of the Extreme Forecast Index (EFI) for 10-metre wind speed in Europe at day 3 and day 5, verified against SYNOP weather station observations. The chart shows 12-month running average values (bold) and seasonal values.

More products for WMO Members

In July 2019, ECMWF substantially increased the amount of weather prediction products it makes available free of charge to Members of the World Meteorological Organization (WMO). All the static web charts and the Ensemble Meteogram on the ECMWF website were made available to all WMO Members.

The additional products enabled a much more comprehensive view than before of atmospheric conditions as predicted by ECMWF, including near-surface weather conditions. The products included both probabilistic and deterministic forecasts.

The changes helped to provide forecasters in the national meteorological and hydrological services (NMHSs) of WMO Members with the information they need to carry out their operational activities.

ECMWF also reduced the cost of the NMHS web non-commercial licence, which gives access to the ecCharts service, and it introduced a cheaper ‘standard’ licence which provides access to a fixed dataset of the licensee’s specification for a reduced fee. The changes were approved by the ECMWF Council at its June 2019 session and are part of the Centre’s efforts to serve WMO Members.

New products for WMO Members
New products for WMO MembersThe new products include static charts such as the Extreme Forecast Index (EFI) for 2 m temperature. The top panel shows the EFI (shading and dashed contours) and the Shift of Tails (SOT) (solid contours) from 00 UTC on 2 October 2019 for 72-hour 2 m mean temperature valid from 8 to 11 October 2019. The bottom panel shows the 99th percentile of the corresponding model climate for those days.

Ocean5 charts made available online

Ocean5 is ECMWF’s current ocean and sea-ice analysis system. It provides initial conditions for the ocean and sea-ice component of ECMWF’s Earth system forecasting system. Ocean5 runs both a behind-real-time stream that produces the Ocean Re-Analysis System 5 (ORAS5) and a near-real-time (NRT) stream. ORAS5 is used for climate monitoring while Ocean5 NRT provides initial conditions for the Centre’s forecasting activities.

In November 2019, charts for each stream were made freely accessible to both internal and external users on the ECMWF website. At the time, the daily sea-surface temperature (SST) maps showed a persistent anomalously warm water mass off the west coast of North America and Alaska. This is illustrated in the figure for 26 November 2019. This ‘marine heatwave’ looked similar to a long-lasting marine heatwave in the same area from 2014 to 2016, baptised ‘the Blob’ by the community.

Sea-surface temperature (SST) anomaly November 2019
Sea-surface temperature (SST) anomaly November 2019This Ocean5 NRT chart shows the SST anomaly on 26 November 2019 computed with respect to the 1993–2016 climate from ORAS5. This kind of chart and many more are freely available to view and download on the ORAS5 and Ocean5 NRT pages on ECMWF’s website.

Testing new extended-range products

A particular concern at the sub-seasonal forecast range (up to 60 days ahead) is the ability to predict high-impact, large-scale and long-lived weather events, such as cold spells or heatwaves. ECMWF’s Strategy to 2025 calls for skilful predictions of regime transitions up to four weeks ahead. In 2019, experimental products to predict cold spells in Europe were made available to registered users. Tests had shown that the products had useful skill up to two and a half weeks ahead.

The products exploit the fact that cold spells in different parts of Europe are closely linked with the occurrence of particular weather regimes. These include a blocking high over Scandinavia (‘Blocking’ in the Figure below), and anomalously high pressure over Greenland combined with low pressure over the Azores (a pattern known as the negative phase of the North Atlantic Oscillation, ‘NAO-’).

ECMWF ensemble forecasts were found to provide reliable probabilities of cold conditions associated with the establishment of the NAO- pattern beyond the medium range. The predictability of such events is enhanced by tropical–extratropical teleconnections. On the other hand, predicting the occurrence of cold events associated with a transition to blocking was found to present a bigger challenge.

Predicting cold spells in the extended forecast range
Predicting cold spells in the extended forecast rangeThe occurrence of cold spells in Europe is closely linked to large-scale atmospheric circulation patterns, such as Blocking and the negative phase of the North Atlantic Oscillation (NAO-). In the figure, this is illustrated for nine different regions in Europe. For each region, the diagrams show the dominant circulation patterns associated with severe cold events (indicated by the dots) that occurred from 1980 to 2015.