ERA5 (Hersbach et al., 2018) is the latest climate reanalysis dataset produced by ECMWF through the Copernicus Climate Change Service (C3S). ERA5 data extending from 1979 to the present became available at the end of 2018, while the second phase, extending back to 1950, is planned for release by autumn 2019. ERA5 is based on the Integrated Forecasting System (IFS) Cycle 41r2 which was operational in 2016. Therefore, ERA5 benefits from a decade of developments in model physics, numerics and data assimilation, compared to the ERA-Interim IFS Cycle 31r2, operational in 2006. In addition to a significantly enhanced horizontal resolution of 31 km, compared to 80 km for ERA-Interim, ERA5 has a number of innovative features, including hourly output, and an uncertainty estimate. ECMWF forecasts from ERA5 analysis show a gain of up to one day in skill with respect to ERA-Interim (Haiden et al., 2018), which is reflected not only in average weather variables but also with regard to large-scale weather patterns such as tropical cyclones.
In this work we used daily maps of precipitation, minimum, mean and maximum surface air temperature, mean sea level pressure, incoming solar radiation, wind speed and relative humidity extracted from ERA5 for 32 complete years between 1986 and 2017. Some variables were then processed to produce estimates of potential evapotranspiration using the Penman-Monteith equation. This represents one of the main dynamic input variables of the hydrological model Lisflood, together with mean surface air temperature and precipitation. ERA5 data was used to model the daily hydrological states over 1986–2017.
ECMWF reforecasts are global scale forecast runs that use the same Integrated Forecasting System (IFS) model version as the real-time ensemble forecasts (ECMWF-ENS), for the past 20 years. Similar to the operational forecasts, reforecasts have horizontal resolutions of 18 km for up to 15-day lead time and 36 km for longer forecast lead times up to 46-day. In this work we used the unperturbed ensemble member (i.e., the control run) of 6-week reforecasts with daily resolution, initialized once per week in the entire period of availability, coupled with the last year of operational forecasts. The collected dataset thus includes 1218 sets of reforecasts with global coverage, spanning 21 years between 21/11/1996 and 25/12/2017. More in details, each of the 21 years of data is composed by 58 weekly reforecasts starting in mid-November of the previous year, so that each day of each year is ultimately simulated in 6 different reforecasts with lead time between 1 and 6 weeks. As for ERA5, we extracted the same set of atmospheric variables and pre-processed them to obtain daily estimates of evapotranspiration, which were used as input of the hydrological model together with mean surface air temperature and precipitation.
Do you have any questions about this protocol?
Post your question to gather feedback from the community. We will also invite the authors of this article to respond.
Tips for asking effective questions
+ Description
Write a detailed description. Include all information that will help others answer your question including experimental processes, conditions, and relevant images.