A new scientific publication in Natural Hazards shows that global river discharge forecasts from the Copernicus Emergency Management Service (CEMS) can be used for reliable seasonal hydrological drought forecasts. The study demonstrates that these forecasts have high skill in anticipating hydrological drought conditions up to six months ahead. The research was conducted by scientists from Politecnico di Milano, the European Centre for Medium-Range Weather Forecasts, and the Joint Research Centre of the European Commission, all partners in the SEED-FD project.
Why Seasonal Drought Forecasts Matter for Anticipatory Action
Drought is often associated with a lack of rainfall, but its most damaging impacts usually appear later, when water stored in soil, rivers, and groundwater becomes scarce. These hydrological droughts occur at the stage when water restrictions are imposed and effects on agriculture, energy supply, and ecosystems become unavoidable. For people and institutions managing water, this is the time at which decisions become critical.
While many early warning systems primarily monitor meteorological drought using rainfall-based indicators, these do not directly describe future water availability in rivers. Hydrological drought focuses on how dry conditions propagate through the water cycle and affect streamflow over time.
Seasonal drought forecasts build on this perspective by looking several months ahead, providing valuable lead time for strategic planning and proactive drought risk management. Unlike flash droughts, which develop rapidly within days or weeks, seasonal hydrological droughts evolve more gradually, making them particularly relevant for early warning and anticipatory action.
A Missing Link in Global Drought Early Warning Systems
At global scale, the Copernicus Emergency Management Service (CEMS) already operates advanced early warning systems. The Global Drought Observatory (GDO) mainly addresses meteorological and agricultural droughts, while the Global Flood Awareness System (GloFAS) provides river discharge forecasts primarily designed for flood monitoring.
Until now, however, no operational system has offered a dedicated seasonal drought forecast for hydrological droughts worldwide. Although GloFAS produces seasonal river discharge forecasts, their potential for predicting drought conditions has not been systematically assessed. The new publication directly addresses this gap.
Key Research Findings on Seasonal Drought Forecasts
The new study conducted by SEED-FD researchers evaluates whether existing global CEMS river discharge forecasts can be used for seasonal hydrological drought forecasting. To this end, river discharge forecasts from the Global Flood Awareness System (GloFAS) were analysed for the period 1991–2022.
Hydrological drought conditions were identified using the Standardized Streamflow Index (SSI), which measures how strongly river discharge deviates from long-term normal conditions. The results show that seasonal drought forecasts perform well not only one month ahead, but also at lead times of three and even six months.
Forecast skill remains high across large parts of the globe, including Europe, Central Africa, southern South America, and East Asia. Importantly, these seasonal drought forecasts consistently outperform a simple “persistence” approach that assumes future conditions remain similar to the present. This highlights the clear added value of these forecasts beyond basic monitoring.
When Drought Forecasts Work Best – and Their Reliability
The study analyses how forecast skill varies with season and region. Seasonal drought forecasts tend to perform best during periods of low average river discharge, when drought impacts are typically most severe. During spring and summer (referring to the Northern Hemisphere), forecast skill is strongly influenced by initial hydrological conditions such as soil moisture. In autumn and winter, the influence of precipitation variability on forecast skill increases.
As a key added value, the study demonstrates that the signal-to-noise ratio (SNR) can be used as a measure of forecast reliability. High SNR values are associated with more reliable forecasts, providing users with transparent information that supports confident decision-making.
Benefits of Seasonal Drought Forecasts for Users
For decision-makers in public authorities, seasonal drought forecasts offer valuable lead time to plan water allocation, manage reservoirs, and prepare drought response strategies. Knowing months in advance that river discharge is likely to be lower than normal supports more proactive and cost-effective action.
Local communities, farmers, and water-dependent businesses also benefit from these seasonal drought forecasts. They allow adjustments in irrigation planning, crop choices, or operational strategies before water scarcity becomes critical. Consultants and intermediaries can use this information to advise clients, support climate risk assessments, and inform longer-term adaptation planning.
By focusing on hydrological droughts, seasonal drought forecasts directly address the stage at which impacts are most strongly experienced. They complement existing meteorological drought information by providing insight into expected changes in water availability, making forecasts more relevant for practical decisions.
Conclusion and Outlook: Seasonal Drought Forecasting in SEED-FD
The results show that a global seasonal drought forecast for hydrological droughts is both feasible and operationally valuable using existing CEMS infrastructure. By systematically evaluating river discharge forecasts for drought conditions, the study lays the groundwork for a hydrologically driven seasonal drought forecast that goes beyond monitoring and supports anticipatory action.
The research was conducted as a direct contribution to the SEED-FD project, whose objectives include expanding the CEMS early warning system portfolio for floods and droughts with a seasonal drought forecast product. The study’s results have already been integrated into the SEED-FD prototype early warning system, bringing together the project’s developments. The next step will focus on validating this prototype using real-world events in hydrologically challenging use case regions, preparing the ground for integration into the operational CEMS system.
The full scientific article, Predicting Hydrological Drought at Global Scale: An Analysis of the CEMS Seasonal Forecasts, is published in the journal Natural Hazards and is available here.
To stay up to date, follow SEED-FD on LinkedIn, Facebook, and X.