In conversation with... Florian Pappenberger

The relationship between weather forecasting and flood prediction is set to become stronger as both activities increasingly rely on an Earth system approach, says Florian Pappenberger, Director of Forecasts at the European Centre for Medium-Range Weather Forecasts (ECMWF).

 Photo credit: ECMWF

Why does weather prediction need an Earth system approach?

Different parts of the natural world, such as the atmosphere, the ocean, sea ice and land, are all interconnected. For example, processes in the ocean and on land influence the atmosphere: a tropical cyclone needs warm water in the ocean, and a heat wave can become more extreme if there is little soil moisture and therefore little evaporation from the land to cool the air.

Conversely, atmospheric conditions such as wind and precipitation influence the state of the ocean and the land. And changes in sea ice can affect the weather, not only in the Arctic, but also in places such as Europe. An Earth system approach takes all these interactions into account. Without it we would be unable to capture these feedbacks, which influence our weather. The best predictions come from viewing the Earth as a single system.


Why use an Earth system approach for flood prediction?

Floods occur because of changes in the distribution and movement of water across the world. Again, such changes are influenced by interconnected developments in the atmosphere, the land, the ocean and water in all its forms. Adopting an Earth system approach means we can make improved and consistent predictions, not only of weather but also of the conditions on the world’s surface, including how much water is flowing in our rivers.

This is particularly true of global predictions, for which we need to model the water cycle across the globe. A single comprehensive Earth system model will be able to make consistent and seamless global weather and flood risk predictions, from hours to seasons ahead.

"Making predictions further into the future will be especially important for droughts"

How does weather prediction support flood and drought forecasting?

In most cases, weather is the main driver for floods and droughts, which are mainly caused by extreme precipitation or prolonged periods without rainfall or snowfall. The further into the future our predictions reach, the more important these drivers become. Better weather forecasts and, in particular, better predictions of the range of possible future precipitation levels, lead to better flood and drought forecasts. As we develop our Earth system modelling capabilities further, we can expect to see even stronger links between global weather forecasts on the one hand and global flood and drought predictions on the other.

How does ECMWF support the EU’s flood prediction capabilities?

ECMWF acts as the computational centre for the EU’s Copernicus Emergency Management Service (CEMS), which produces daily flood and fire risk forecasts. Our world-leading global weather forecasts feed into these flood and fire risk prediction systems. In the case of floods, these forecasts are used in the European Flood Awareness System (EFAS) and the Global Flood Awareness System (GloFAS), which provide early information on high river flows around the world.

We are also involved in a range of EU-funded flood-related research initiatives and support the South- East European Multi-Hazard Early Warning Advisory System. This is an initiative of the World Meteorological Organization and south-east European countries supported by the EU.

What kind of progress can we expect to see in European flood and drought prediction over the next ten years?

Progress in Earth system modelling will reduce or remove the need for separate systems to predict the weather and floods or droughts. At the same time, the spatial resolution of meteorological and hydrological forecasts will increase, and our representation of Earth system processes will improve. We will also get better at representing uncertainties. Today we struggle to predict floods that are generated in small, steep valleys or across small connective systems.

Not only will we be able to better predict such floods, we will also be able to predict them further in advance. Making predictions further into the future will be especially important for droughts; if we improve the prediction of temperature and precipitation up to seasons ahead, we will get a better handle on droughts. Such forecasts could, for example, support planning in agriculture: the type of crops we plant, the best use of water for irrigation and the optimum harvest time. This will improve productivity as well as food security.

"A single comprehensive Earth system model will be able to make consistent and seamless global weather and flood risk predictions, from hours to seasons ahead"

The theme of this year’s world water day is the link between water and climate change. How can ECMWF data help us to understand that link?

We have produced a wealth of data to investigate the links between water and climate change. For example, we have combined data from CEMS and the Copernicus Climate Change Service (C3S) to produce a 40-year record of globally consistent river flows, together with a globally consistent record of the atmosphere and the land surface. Studying this record will allow us to understand how extremes in a changing climate are linked.

For example, it will tell us through which processes tropical cyclones are linked to extreme flooding. As a partner in the EU-funded Horizon 2020 CAFE project (Climate Advanced Forecasting of sub-seasonal Extremes), we are studying the large-scale processes that can make heatwaves and droughts predictable up to weeks ahead. A better understanding of the links between water and climate change will give us a better understanding of climate-related risks.

Read the most recent articles written by The Parliament Magazine - 5 Questions with... Pierre Karleskind

Share this page