ECMWF `ERA5`

In this file, the details of the dataset is explained.

Location of Dataset Files

The global ERA5 dataset is located under the following directory accessible from Compute Canada (CC) Graham Cluster:

# DRAC Graham HPC location
/project/rpp-kshook/CompHydCore/climateForcingData/ERA5/ERA5_for_SUMMA/2_merged_data # rpp-kshook allocation

# UCalgary ARC HPC location
/work/comphyd_lab/data/meteorological-data/era5 # comphyd_lab location

# Perdue University Anvil HPC location
/anvil/datasets/meteorological/ERA5_small # Anvil community storage

and the structure of the dataset hourly files is as following:

/path/to/dataset/dir/
├── ERA5_merged_195001.nc
├── ERA5_merged_%y%m.nc
├── .
├── .
├── .
└── ERA5_merged_202012.nc

Coordinate Variables and Time-stamps

Coordinate Variables

The coordinate variables of the ERA5 simulations are longitude and latitude representing the longitude and latitude points, respectively. ### Time-stamps The time-stamps are included in the original files.

Dataset Variables

The NetCDF files of the dataset contain 7 variables needed to force hydrological models. You may see a list of variables by using the ncdump -h command on one of the files:

foo@bar:~$ module load cdo/2.0.4
foo@bar:~$ module load nco/5.0.6
foo@bar:~$ ncdump -h  /project/rpp-kshook/CompHydCore/climateForcingData/ERA5/ERA5_for_SUMMA/2_merged_data/ERA5_merged_195001.nc

Spatial Extent

The spatial extent of the ERA5 is on latitutes from +90 to -90 and longitudes from -180 to 179.75. The resolution is 0.25 degrees.

Temporal Extent

The time-steps are hourly covering from January 1950 to December 2020.

Short Description on `ERA5` Variables

In most hydrological modelling applications, usually 7 variables are needed detailed as following: 1) specific humidity at 2 meters, 2) surface pressure, 3) air temperature at 2 meters, 4) wind speed at 10 meters, 5) precipitation, 6) downward short wave radiation, and 7) downward long wave radiation. These variables are available through the current ERA5 dataset and their details are described in the table below:

ERA5 Variables
Variable Name	ERA5 Variable	Original Shortname	Parameter ID	Unit	IPCC abbreviation	Comments
surface pressure	`airpres`	sp	134	Pa	ps
specific humidity @2m	`spechum`	q	133	kg/kg	huss
air temperature	`airtemp`	t	130	k	tas
wind speed @10m	`windspd`	u,v	131,132	m/s	wspd	WIND=SQRT(u^2+v^2)
precipitation	`pptrate`	mtpr	235055	mm/hr		mean total precipitation rate
short wave radiation	`SWRadAtm`	msdwswrf	235035	W m-2	rsds``
long wave radiation	`LWRadAtm`	msdwlwrd	235036	W m-2	rlds``

For a complete catalog of the dataset, see here.

Downloading Original ERA5 Data

The scripts to download the original ERA5 data are contained in the `ERA5_downloads.zip <./ERA5_downloads.zip>`__ file located in the current directory of the repository. The scripts are written by Quoqiang Tang and Wouter Knoben as part of the Community Workflows to Advance Reproducibility in Hydrological Modelling (CWARHM). “ERA5 data preparation includes interactions between an atmospheric model and a land surface model. … [Original] ERA5 data is available at 137 different pressure levels (i.e. some height above the surface), as well as at the surface. The lowest atmospheric level is L137, at geopotential and geometric altitude 10m and data here relies only on the atmospheric model. Any variables at a height lower than L137 (i.e., at the surface) are the result of interpolation between atmospheric model and land model. We want to use only the outcomes from the ECMWF atmospheric model […]. Therefore, we obtain (1) air temperature, (2) wind speed and (3) specific humidity at the lowest pressure level (L137), (4) Precipitation, (5) downward shortwave radiation, (6) downward longwave radiation and (7) air pressure are unaffected by the land model coupling and can be downloaded at the surface level. This is beneficial because surface-level downloads are substantially faster than pressure-level downloads. [^1]” [^1]: from: https://github.com/CH-Earth/CWARHM/tree/main/3a_forcing#forcing-needed-to-run-summa

ECMWF ERA5