Forecasting of hydrometeorological drought in Europe months in advance

Contributed by: Samuel J. Sutanto and Henny A.J. Van Lanen

Hydrology and Quantitative Water Management Group, Wageningen University and Research, the Netherlands


In the past three years, from 2018 to 2020, extensive areas in Europe once more suffered from extreme drought. The economic and environmental losses caused by drought are enormous and have affected many sectors, such as agriculture and livestock farming, water transportation, energy and industry and aquatic ecosystems. To reduce losses, a Drought Early Warning System (DEWS) would offer great help in providing the government, water managers, and end-users with information of ongoing and foreseen drought events. In Europe, the European Drought Observatory (EDO) has been established and operational for about a decade (Fig. 1). This web-based platform provides warning based on combined drought indices (the Standardized Precipitation Index (SPI), soil moisture anomaly, and vegetation condition anomaly) ( Recently, drought forecasting has been implemented in EDO, which includes meteorological drought forecasts based on SPI for 3 months ahead and soil moisture anomaly (SMI) for 7 days ahead. EDO is still limited in hydrological drought predictions, that is, forecasted droughts in groundwater and river flow are not provided yet. Lately, a DEWS to complement the EDO has been developed within the EU-funded H2020 ANYWHERE project (, i.e. the ANYWHERE DEWS (ADEWS). This system is a part of Multi-Hazard EWS (MH-EWS) that predicts major weather-related natural hazards in Europe, such as floods, flash floods, debris flows, landslides, storm surges, heatwaves and air quality, wildfires, droughts, convective storms, severe winds, and heavy snowfall (for all products see: The ANYWHERE DEWS (ADEWS) forecasts drought in different water cycle components (drought in precipitation, soil moisture, runoff[1], groundwater, and discharge[2]) using both the threshold and standardized approaches with lead times up to 7 months (see Sutanto et al., 2020a for detailed information).

Figure 1 The EDO EWS (left) and the ANYWHERE MH-EWS (right). The EDO panel provides the Combined Drought Index (CDI) derived from observations for the period 11-20 May 2020. The MH-EWS panel shows the weather-related hazards provided by the MH-EWS (black bar) and the map with the forecasted SPEI-3 for 1 July 2020 initiated early May 2020.

The capability of ADEWS to predict drought a few months in advance was demonstrated for the 2003 and 2018 droughts (Sutanto et al., 2020a; 2020b). In this context, among others, the evolution of meteorological drought (SPI-6 and the Standardized Precipitation Evaporation Index, SPEI-6), and the hydrological drought (the Standardized Runoff Index (SRI-6) and the Standardized Groundwater Index, (SGI-1)) were forecasted at the beginning of May 2018 for months May (lead time, LT, 1 month), June (LT 2 months), and July (LT 3 months) (Sutanto et al., 2020a). For the 2020 drought, we show the forecasted drought hazard initiated in May 2020 for LT = 1 and 3 months (May and July, respectively, Fig. 2). The meteorological drought that occurred in the spring of 2020 in the west, central, and east Europe (Fig. 2a) was foreseen by the ADEWS in May 2020. The meteorological drought severity was forecasted to decrease in July (Fig. 2b). On the other hand, the ADEWS predicted more extreme hydrological drought (drought in runoff) than meteorological drought in May 2020, especially in the north and east Europe (Fig. 2c). Unlike the meteorological drought, the hydrological drought condition in July is predicted to remain the same as in May in these regions, except in north Europe (Fig. 2d). Many rivers in Europe (except in the south) were forecasted to experience drought from mild to extreme (Fig. 2e). Similar to drought in runoff, rivers in northern Europe would have more water in the summer (July) (Fig. 2f).

Figure 2 Forecasted drought in precipitation: a) SPI-3, LT=1, b) SPI-3, LT=3; in runoff: c) SRI-1, LT=1, d) SRI-1, LT=3; in streamflow: e) SSI-1, LT=1, and f) SSI-1, LT=3 using forecasts issued at the beginning of May 2020.  Acronyms at the bottom-right panel (Fig. 2e, 2f) stand for: Extreme Drought (ED), Severe Drought (SD), Moderate Drought (MoD), Mild Drought (MD), Normal (N), Mild Wet (MW), Moderate Wet (MoD), Very Wet (VW), and Extreme Wet (EW).

Although the ADEWS showed to have potential to predict droughts in 2003, 2018, and 2020 relatively well, we found that the drought forecasts commonly underestimated drought severity (e.g. drought in precipitation and runoff) in west and central Europe compared to observation (Sutanto et al., 2020a; 2020b). The underestimation of drought severity in these regions also occurred for forecasted drought in 2020. Nevertheless, the forecasts clearly showed the upcoming drought event to some extent. Our previous studies on drought forecast skill showed that the ADEWS can predict the drought hazard more or less up to 2-3 months in advance (Sutanto et al., 2019a; 2020b; Van Hateren et al., 2019).

The ADEWS has been in a pre-operational mode since spring 2018 and has been online (7/24) for 2 years (until the end of July 2020). The ADEWS is ready to become operational across Europe and to be further developed when resources become available. Using this system, drought hazard can be forecasted some months ahead. Still missing in the system is drought impact functions for the whole of Europe to forecast drought impacts, which is possible with the help of the European Drought Impact Inventory (EDII) database and machine learning techniques (e.g., Sutanto et al., 2019a). The ADEWS is a part of MH-EWS, which provides different weather-related hazard forecasts. However, the MH-EWS only forecasts hazards as a single hazard, and no compound and cascading hazards are forecasted (only different hazards can be overlaid). The occurrence of compound and cascading hazards usually will lead to a higher impact than a single hazard alone. The first steps of the development of the MH-EWS that provides information on the prediction of compound and cascading events have been taken (Vitolo et al., 2019; Sutanto et al., 2019b) and must be considered for future work.


  • Sutanto, S.J., Van Lanen, H.A.J., Wetterhall, F., and Llort, X.: Potential of pan-European seasonal hydro-meteorological drought forecasts obtained from a Multi-Hazard Early Warning System, BAMS, 101, 368–393,, 2020a.
  • Sutanto, S.J., Wetterhall, F., and Van Lanen, H.A.J.: Hydrological drought forecasts outperform meteorological drought forecasts, Environ. Res. Lett., 15, 084010,, 2020b.
  • Sutanto, S.J., Van der Weert, M., Wanders, N., Blauhut, V., and Van Lanen, H.A.J.: Moving from drought hazard to impact forecasts, Nature Communications, 10:495,, 2019a.
  • Van Hateren, T.C., Sutanto, S.J., and Van Lanen, H.A.J.: Evaluating skill and robustness of seasonal meteorological and hydrological drought forecasts at the catchment scale – case catalonia (spain), Environment International, 133, 105206,, 2019.
  • Vitolo, C., Di Napoli, C., Di Giuseppe, F., Cloke, H.L., and Pappenberger, F.: mapping combined wildfire and heat stress hazards to improve evidence-based decision making, Environment International, 127, 21–34,, 2019.
  • Sutanto, S.J., Vitolo, C., Di Napoli, C., D’Andrea, M., and Van Lanen, H.A.J.: Heatwaves, droughts, and fires: exploring compound and cascading dry hazards at the pan-European scale, Environment International, 134, 105276,, 2019b.

[1] Runoff (m3/s/m2), sometimes called specific discharge, is the volume of water that flows through the subsurface (mainly aquifers) and occasionally over the surface towards the surface water system (e.g. rivers). Runoff generation is commonly implemented in large-scale gridded models (e.g. Global Hydrological Models, Land Surface Models)

[2] Discharge (m3/s) is the volume of water flowing through a river, which is measured in a gauging station or simulated in a river node or outlet of a catchment. Large-scale gridded models may route the generated runoff from the upstream (contributing) area towards a nodal point at the river network to obtain the discharge.

Leave a Comment