EU Water Stress: FR/BE situation & Copernicus ECMWF ensemble forecast

04 May 2019

Read on Twitter:


Water deficit affecting agriculture in France’s @Departement63 Puy-de-Dome @FannyGuine

Viz below ↓

@CopernicusEMS Combined #Drought Indicator (CDI) - to indicate agricultural soil drought [] - learns:

Situation in France

Soil drought in @auvergnerhalpes in line with March DREAL analysis []:

Globally depressed streamflow: a majority of test points has an average flow rate below the ‘Five-Year Dry Flow’ [= monthly flow with a 4/5 probability of being exceeded each year. It makes it possible to characterize a calendar month of low hydraulicity]

Majority of groundwater in the western part of the @auvergnerhalpes region has a very low level [red / orange dots]. With the recovery of vegetation, only very heavy spring rains are likely to reverse current trends.

In the @auvergnerhalpes region, 2 departments are vigilant for water use (Ain and Rhône - since April 1).

In France, 10 departments have a water usage restriction (beyond vigilance):

Situation in Belgium

In Belgium (Flanders), a similar deficit in groundwater level is observed: 1/3 of the measurement points are very low [brown], 38% of the measuring points have a normal level [white]. @DOVdovVO A debate is ongoing on water restrictions to avoid shortages during the summer.

Situation in the EU

In the EU, @EUEnvironment [latest data 2015,] :

Future projection of soil water content

How will the soil water content evolve, depending on future #emissions? #SWICCA @CopernicusECMWF



France (Auvergne Rhone-Alpes)

For the @auvergnerhalpes water catchment area, a decreasing soil water content in a high emission scenario (#RCP 8.5) can be observed as well [both for 2050 and 2080] @Hydro_research

Belgium (Flanders)

In Flanders, the effect of future emissions on soil water content is less outspoken according to @CopernicusECMWF ensembles, although also indicating lower soil water content - specifically during summer - in a high emission scenario (2080) @klimaat_be @DOVdovVO @MarjoVanoppen


@Hydro_research, feel free to comment/evaluate these conclusions! Some thoughts: 1) For the long-term forecast, there is a 50km raster available for the #HYPE, #VIC421 and #LISFLOOD models, but for the catchment simulation (as above), only the #HYPE model is used. Why is this?

2) [a] For the short-term seasonal forecast [ precip., T and river flow], only the #HYPE model can be used and [b] soil water content simulation is not available. Could you explain the reasons for those (a and b) ?