EFAS forecasts are generated as a combination of Numerical Weather Prediction meteorological forecasts input and a hydrological model. Only EFAS medium-range forecasts, issued twice daily at 00:00UTC and 12:00UTC, are currently made available to the public as data files. Their configurations are detailed in table 1 below (for more details on the meteorological forcings please see https://www.efas.eu/en/meteorological-forecasts ).

Table 1: EFAS medium-range forecast configuration

EFAS Historical simulation forced with observations

EFAS historical simulation forced with observations is a long (at least 20 years) hydrological time series simulation.  It is considered as a reference simulation for the system, and is generated every time a new EFAS operational configuration is released. It is used to calculate statistics against which to compare the forecasts, such as river flood magnitude for given return periods, or to verify the performance of the forecasts.

EFAS historical simulation forced with observations (SFO) is produced by forcing the EFAS operational modelling chain with consolidated gridded field of observed meteorological variables (precipitation and temperature) over a time period as long as possible; e.g. for EFAS3.0 the EFAS historical simulation spans 1990-2016.

From 2017 onward, simulations forced with observations make use of gridded field of observed meteorological variables from the operational EFAS and are identified as EFAS operational simulations forced with observations (operational SFO).

Users who want to full SFO time series are advised to download both historical SFO and operational SFO and merge them together.

As the initial conditions used to initialise EFAS historical SFO for the first day of the simulation correspond to averaged initial values (the model is so called 'cold-started'), EFAS historical simulation forced with observations is available from 1 January 1991, when the initial conditions don't influence the simulation results.

EFAS variable river discharge (dis) is stored as average over the time step of the model (6-hourly or 24-hourly), with time stamp (date and time of end of averaging period) explicit in the GRIB short name.

In EFAS, river discharge is calculated in each grid box of the hydrological model. However, not all grid boxes correspond to a river segment, and because of the relatively coarse resolution of the EFAS gridded domain, the modelled river network in EFAS might be associated with slightly different latitude-longitude than in the real world. It is hence essential that when extracting river discharge data, the user also checks the upstream area of the grid box of interest. The EFAS upstream area can be found in the Auxiliary Data pages.

Some short scripts to use as examples are provided in the Working with CEMS-Flood Data pages. 

Snow depth water equivalent represents the amount of water stored as ice and snow given as an instantaneous value at the end of each model time step. The time stamp (date and time of end of time step period) is explicit in the GRIB short name.

In EFAS, snow depth water equivalent in each grid box is calculated from three elevation-related bands (a, b and c) to account for the difference in the melting rate with snow condition. However, only the more directly usable composite snow map representing the total snow water equivalent for each grid box is made available.

Some short scripts to use as examples are provided in the Working with CEMS-Flood Data pages. 

Soil moisture content represents the amount of water stored in the soil layers given as an instantaneous value at the end of each model time step. The time stamp (date and time of end of time step period) is explicit in the GRIB short name.

EFAS soil moisture content in each grid box is calculated in three layers, top (1), intermediate (2) and bottom (3), each associated with a given soil depth (depth from top to bottom of that layer). Information for each 3 layers is provided as separate data field, the soil depth.

EFAS soil moisture information is stored as volumetric soil moisture content for each layer. The associated volume of water for that layer is volumetric soil moisture content multiplied by the layer thickness. The thickness of a specific layer is the difference between the depth of two consecutive layers.

The total volume soil moisture content is the sum of volume of water for each layer. Note that in EFAS-IS, only soil moisture of the first 2 layers is shown.

Some short scripts to use as examples are provided in the Working with CEMS-Flood Data pages. 

This dataset contains global modelled daily data of river discharge from GloFAS. River discharge. This dataset is simulated by forcing the the LISFLOOD hydrological modelling chain with inputs from the ERA5 global reanalysis. Data availability for the historical simulation is from 1979-01-01 up to near real time. Data can be accessed is several ways, the most common is through the CDS (see Data Access).

Table X: River discharge and related historical data from GloFAS

This dataset contains global modelled daily data of river discharge forced with meteorological forecasts from GloFAS. This dataset is simulated by forcing the the LISFLOOD hydrological modelling chain with input from ECMWF ensemble forecast combined with the ECMWF extended-range ensemble forecast up to 30 days. Data availability for the GloFAS forecast is from 2019-11-05 up to near real time. Data can be accessed is several ways, the most common is through the CDS (see Data Access).

Table X: River discharge and related forecasted data from GloFAS

This dataset provides a gridded modelled time series of river discharge, forced with medium- to sub-seasonal range meteorological reforecasts from GloFAS. This dataset was produced by forcing the LISFLOOD hydrological modelling chain with input from the ECMWF-ENS 11-member ensemble reforecasts. Reforecasts are forecasts run over past dates, and those presented here are used for providing a suitably long time period against which the skill of the 30-day real-time operational forecast can be assessed. The reforecasts are initialised twice weekly with lead times up to 46 days, at 24-hour steps for 20 years in the recent history. For more specific information on the how the reforecast dataset is produced we refer to the documentation. Data can be accessed is several ways, the most common is through the CDS (see Data Access).

Table X: Reforecasts of river discharge and related data from GloFAS

This dataset provides a gridded modelled time series of river discharge, forced with seasonal range meteorological forecasts from GloFAS. This dataset was produced by forcing the LISFLOOD hydrological model at a 0.1° (~11 km at the equator) resolution with downscaled runoff forecasts from the ECMWF 51-member ensemble seasonal forecasting system, SEAS5. The forecasts are initialised on the first of each month with a 24-hourly time step, and cover 123 days. Data can be accessed is several ways, the most common is through the CDS (see Data Access).

Table X: Seasonal river discharge forecasts andSeasonal forecasts of river discharge and related data from GloFAS

This dataset provides a gridded modelled time series of river discharge, forced with seasonal range meteorological forecasts from GloFAS. This dataset was produced by forcing the LISFLOOD hydrological model at a 0.1° (~11 km at the equator) resolution with downscaled runoff forecasts from the ECMWF 51-member ensemble seasonal forecasting system, SEAS5. The forecasts are initialised on the first of each month with a 24-hourly time step, and cover 123 days. Data can be accessed is several ways, the most common is through the CDS (see Data Access).

Table X: Seasonal reforecasts of river discharge and related data from GloFAS

GloFAS variable river discharge (dis) is stored as average over the time step of the model (24-hourly), with time stamp (date and time of end of averaging period) explicit in the GRIB short name.

In GloFAS, river discharge is calculated in each grid box of the hydrological model. However, not all grid boxes correspond to a river segment, and because of the relatively coarse resolution of the GloFAS gridded domain, the modelled river network in GloFAS might be associated with slightly different latitude-longitude than in the real world. It is hence essential that when extracting river discharge data, the user also checks the upstream area of the grid box of interest. The GloFAS upstream area can be found in the Auxiliary Data pages.

Some short scripts to use as examples are provided in the Working with CEMS-Flood Data pages.