EP flux data overview
Here you will find several EP flux data sets calculated from ECMWF ERA-Interim reanalysis (Simmons et al., 2006), ECMWF ERA-40
reanalysis (Uppala et al., 2005), ECMWF operational analyses, NCEP reanalysis (Kistler et al., 2001) and Japanese reanalysis
(Onogi et al., 2007). If these data are used in scientific publications, please let us know and acknowledge properly. If you have
further questions, mail us.
Cautionary note: Use the data at your own risk. There are several issues in the used data sets, e.g. inhomogeneities caused by the
introduction of satellite data into the analyses in 1979 and a poorer quality of the data in the southern hemisphere.
Calculation of EP flux
EP fluxes are calculated according to equations 3.5.3a and b of Andrews et al. (1987). Derivatives are approximated by centered differences.
The term including the vertical wind is not calculated, since the quality of vertical wind data in the analyses is not sufficient.
Used constants: p0=1000 hPa, T0=240 K, R=287 J/(kg K), g0=9.80655 m2/s, a=6371 km, angular frequency Earth 7.292e-5 1/s.
EP flux through the tropopause
We provide values of NCEP, JRA, ECMWF ERA-40, ECMWF ERA-Interim and ECMWF operational analysis EP fluxes (vertical component)
through the tropopause at 100 hPa and 45-75 degrees north or south (NCEP 01/1948-12/2012, JRA 11/1978-12/2004, ECMWF ERA-40
09/1957-08/2002, ECMWF ERA-Interim 01/1979-12/2012, ECMWF operational 01/1998-12/2012) for every 6 hour time step of the analyses.
Unit is 105 kg/s2. The file contains the columns "Year", "Month", "Day", "Hour", "EP flux north" and "EP flux south".
Here are the same data as monthly means:
Complete gridded EP flux data set
These are complete gridded EP flux data sets for every latitude, pressure level and date of the used analysis, both for the horizontal
and vertical component of the EP flux vector. Data is available for NCEP reanalysis (01/1948-12/2012), ECMWF ERA-Interim reanalysis
(01/1979-12/2012), ECMWF ERA-40 reanalysis (09/1957-08/2002), ECMWF operational analysis (01/1998-12/2012) and JRA reanalysis
(11/1978-12/2004). Additionally, we provide monthly means for every latitude and pressure level. Files are in netcdf format.
Files are self-explanatory (look into the header, e.g. with ncdump -h). The format was changed from ASCII files to netcdf due to
the large amount of data. There are netcdf libraries for almost all programming languages and scientific software, please have a
look on the internet.
(*) Please note: The number of vertical levels in ECMWF operational data changes in dependence of the date in the analysis data.
The vertical levels given in the files are a minimal subset of the vertical levels originally in the analysis which are available
for every date.
ECMWF ERA-Interim Gridded Sixhourly
ECMWF ERA-40 EP flux Gridded Sixhourly
ECMWF Operational Gridded Sixhourly (*)
NCEP EP flux Gridded Sixhourly
JRA EP flux Gridded Sixhourly
EP flux proxy for use in regression models
This is a proxy for use in multiple linear regression models that want to use EP flux as a variable. It is only intended for data time series of a certain month/season (e.g. an ozone column time series that contains data from January 1970, January 1971 and so on). It is not possible to use this proxy for data including all months of the year simultaneously. Unit is 105 kg/s2. The files contain the columns "Year", "Month" and "Proxy". The proxy is the vertical component of the EP flux through the tropopause (i.e. 100 hPa) in mid-latitudes, and is averaged appropriately to match the influence on ozone transport correctly (see Andrews et al., 1987 and Wohltmann et al., 2007 for more information).
The proxy for the northern hemisphere (use for latitudes >30 degree N) is averaged temporally from the last preceding 1st of October to the 15th of the "Month" given in the file for the months from November to March. It is averaged from 1st of October to 15th of April for the months from April to September. The proxy is undefined for October. It is averaged spatially from 45-75 degrees N and at 100 hPa.
The proxy for the southern hemisphere (use for latitudes >30 degree S) is averaged from 1st of April to the 15th of the "Month" given in the file for the months from May to September. It is averaged from 1st of April to the 15th of October for the months between October to March. It is undefined for April. It is averaged spatially from 45-75 degrees S and at 100 hPa.
The proxy for the tropics (use for latitudes lower than 30 degrees) is averaged over the 3 months before the "Month" given in the file and the 15th of the"Month" given in the file. It is averaged from 45-75 degrees S for May to October and from 45-75 degrees N for the remaing months, all at 100 hPa.
All proxies are averaged according to Wohltmann et al., 2007 (this has changed compared to the old web page).
I. Wohltmann et al. (2007), A process-oriented regression model for column ozone, J. Geophys. Res., 112, D12304, doi:10.1029/2006JD007573.
A. Simmons et al. (2006), ERA Interim: New ECMWF reanalysis products from 1989 onwards, ECMWF News Letter, 110, 25-35.
S. M. Uppala et al. (2005), The ERA-40 reanalysis, Q. J. R. Meteorol. Soc., 131, 2961-3012, 2005.
R. E. Kistler et al. (2001), The NCEP-NCAR 50-year reanalysis: Monthly means CD-ROM and documentation, Bull. Amer. Meteorol. Soc., 82, 247-268.
K. Onogi et al. (2007), The JRA-25 Reanalysis, J. Meteor. Soc. Japan,.85, No. 3, 369-432.
D. G. Andrews, J. R. Holton and C. B. Leovy (1987), Middle Atmosphere Dynamics, Academic Press.
A. C. Fusco and M. L. Salby (1999), Interannual variations of total ozone and their relationship to variations of planetary wave activity, J. Clim., 12, 6, 1619-1629, & Corrigendum 12, 10, 3165.
W. J. Randel, F. Wu and R. Stolarski (2002), Changes in column ozone correlated with the stratospheric EP flux, J. Meteor. Soc. Jap., 80, 4b, 849-862.