EN4¶

EN4: Quality-Controlled Subsurface Ocean Temperature and Salinity Profiles and Objective Analyses, version EN.4.2.2

Available

At a glance¶

Note on period: On 2026-10-12 the latest preliminary files on the download page were for August 2026 (202608), and the yearly zip files run through 2026. Data are available about two months after collection, with preliminary versions released about a month earlier. Early decades (especially before 1950) have very sparse observations, so the analyses stay close to climatology.

Overview¶

EN4 is the Met Office Hadley Centre's global subsurface ocean temperature and salinity dataset. It consists of two products: observed profiles with data-quality information, and monthly 1° objective analyses formed from those profiles, with uncertainty estimates. Profiles come from Argo, ASBO, GTSPP and the World Ocean Database, and the record runs from 1900 to the present. Four versions apply different XBT and MBT bias-correction schemes (g10, l09, c13, c14), giving a measure of structural uncertainty. EN4 is widely used for ocean heat and salinity change studies, model evaluation and comparison with ocean reanalyses.

Available variables¶

Based on the EN.4.2.2 file-format pages, FAQ and Product User Guide. Analysis temperature is potential temperature (reference pressure 0 dbar) stored in Kelvin; salinity is practical salinity with units of "1" (changed from psu in EN.4.2.0 for CF compliance, values unchanged).

instantaneous mean accumulated maximum minimum static — instantaneous: value at the time stamp · mean/accumulated/maximum/minimum: over the indicated interval · static: does not change with time

Objective analyses (EN.4.2.2.f.analysis..YYYYMM.nc)¶

Profiles (EN.4.2.2.f.profiles..YYYYMM.nc)¶

The full variable list is in the official documentation.

How to get the data¶

• Met Office Hadley Centre — EN.4.2.2 download page — Web download (yearly zip files; recent preliminary months as .nc.gz) / wget list files No login needed
  • Profiles and analyses are provided separately for each bias-correction option (g10, l09, c13, c14). For multi-year downloads, use the per-option list files, e.g. EN.4.2.2.analyses.g10.download-list.txt, with wget -i. No account is needed. Data without bias corrections are available on request.
• Earlier versions (EN.4.2.1 and before) — Web download No login needed
  • No longer updated; use only to reproduce earlier studies.

Example code¶

Download EN.4.2.2.analyses.g10.2020.zip from the download page and unzip it first; it contains the monthly files EN.4.2.2.f.analysis.g10.2020MM.nc.

# pip install xarray netCDF4 dask
import xarray as xr

ds = xr.open_mfdataset("EN.4.2.2.f.analysis.g10.2020*.nc")
temp_c = ds["temperature"] - 273.15          # Kelvin -> degC (potential temperature)
w = ds["temperature_observation_weights"]

# Mask grid boxes where observations had little influence (mostly background/climatology)
temp_obs = temp_c.where(w > 0.5)
print(temp_obs.sel(depth=5, method="nearest").sel(lat=slice(30, 45), lon=slice(120, 135)).mean(["lat", "lon"]).values)

Practical tips¶

• Where observations are missing, the analyses relax back to the 1971–2000 climatology, so variability and trends are damped in sparsely sampled periods and regions (before 1950, high southern latitudes, the deep ocean). The FAQ discourages trend analysis where observation weights are low; map *_observation_weights before computing heat-content trends.
• g10, l09, c13 and c14 differ only in their XBT/MBT bias corrections (g10: Gouretski & Reseghetti 2010 XBT + Gouretski & Cheng 2020 MBT; l09: Levitus et al. 2009 XBT and MBT; c13: Cowley et al. 2013 XBT + Levitus et al. 2009 MBT; c14: Cheng et al. 2014 XBT + Gouretski & Cheng 2020 MBT). The Met Office recommends looking at all four to see what impact the choice makes. Quote the version (EN.4.2.2) and the correction used in publications.
• Analysis temperature is potential temperature in Kelvin (with the 273.15 offset in the file); profile POTM_CORRECTED is in °C. Do not mix the two.
• Where no background salinity is available, potential temperature is computed assuming a salinity of 35.0 and flagged as untrustworthy. Always apply the QC flags when using the profiles directly.
• World Ocean Database (WOD18) data were taken as a single download in October 2020 and will not be updated until the next EN4 version, so there are no WOD-sourced observations after late 2020. Recent years rely mainly on Argo and GTSPP; if coastal or ship-based observations matter for your study, check WOD as well.
• EN4 does not take data from surface-only sources. For gridded sea surface temperature, the FAQ recommends HadSST4 or HadISST.

Notices and change history¶

References for your paper¶

Reference papers¶

EN4 reference paper (required) — DOI 10.1002/2013JC009067

@article{good2013en4,
  author  = {Good, Simon A. and Martin, Matthew J. and Rayner, Nick A.},
  title   = {{EN4}: Quality controlled ocean temperature and salinity profiles and monthly objective analyses with uncertainty estimates},
  journal = {Journal of Geophysical Research: Oceans},
  volume  = {118},
  number  = {12},
  pages   = {6704--6716},
  year    = {2013},
  doi     = {10.1002/2013JC009067}
}

g10 XBT correction (Gouretski & Reseghetti 2010) — DOI 10.1016/j.dsr.2010.03.011

@article{gouretski2010xbt,
  author  = {Gouretski, Viktor and Reseghetti, Franco},
  title   = {On depth and temperature biases in bathythermograph data: Development of a new correction scheme based on analysis of a global ocean database},
  journal = {Deep Sea Research Part I: Oceanographic Research Papers},
  volume  = {57},
  number  = {6},
  pages   = {812--833},
  year    = {2010},
  doi     = {10.1016/j.dsr.2010.03.011}
}

l09 XBT and MBT corrections (Levitus et al. 2009) — DOI 10.1029/2008GL037155

@article{levitus2009ohc,
  author  = {Levitus, S. and Antonov, J. I. and Boyer, T. P. and Locarnini, R. A. and Garcia, H. E. and Mishonov, A. V.},
  title   = {Global ocean heat content 1955--2008 in light of recently revealed instrumentation problems},
  journal = {Geophysical Research Letters},
  volume  = {36},
  number  = {7},
  pages   = {L07608},
  year    = {2009},
  doi     = {10.1029/2008GL037155}
}

c13 XBT correction (Cowley et al. 2013) — DOI 10.1175/JTECH-D-12-00127.1

@article{cowley2013xbt,
  author  = {Cowley, Rebecca and Wijffels, Susan and Cheng, Lijing and Boyer, Tim and Kizu, Shoichi},
  title   = {Biases in Expendable Bathythermograph Data: A New View Based on Historical Side-by-Side Comparisons},
  journal = {Journal of Atmospheric and Oceanic Technology},
  volume  = {30},
  number  = {6},
  pages   = {1195--1225},
  year    = {2013},
  doi     = {10.1175/JTECH-D-12-00127.1}
}

c14 XBT correction (Cheng et al. 2014) — DOI 10.1175/JTECH-D-13-00197.1

@article{cheng2014xbt,
  author  = {Cheng, Lijing and Zhu, Jiang and Cowley, Rebecca and Boyer, Tim and Wijffels, Susan},
  title   = {Time, Probe Type, and Temperature Variable Bias Corrections to Historical Expendable Bathythermograph Observations},
  journal = {Journal of Atmospheric and Oceanic Technology},
  volume  = {31},
  number  = {8},
  pages   = {1793--1825},
  year    = {2014},
  doi     = {10.1175/JTECH-D-13-00197.1}
}

g10 and c14 MBT correction (Gouretski & Cheng 2020) — DOI 10.1175/JTECH-D-19-0205.1

@article{gouretski2020mbt,
  author  = {Gouretski, Viktor and Cheng, Lijing},
  title   = {Correction for Systematic Errors in the Global Dataset of Temperature Profiles from Mechanical Bathythermographs},
  journal = {Journal of Atmospheric and Oceanic Technology},
  volume  = {37},
  number  = {5},
  pages   = {841--855},
  year    = {2020},
  doi     = {10.1175/JTECH-D-19-0205.1}
}

Acknowledgements¶

The terms and conditions require a credit line: "EN.4.2.2 data were obtained from https://www.metoffice.gov.uk/hadobs/en4/ and are © Crown Copyright, Met Office, [year of first publication], provided under a Non-Commercial Government Licence" (outside the UK, "© British Crown Copyright"). Quote the version, the bias corrections used and the download date. Cite Good et al. (2013) and the paper(s) for the bias corrections you used, and acknowledge the input data sources (Argo, ASBO, GTSPP, WOD18) as set out in the Product User Guide. The guide also lists the Hanawa et al. (1995) and Kizu et al. (2005) XBT fall-rate adjustments, applied in all versions, as references.

Information on the datasets used in this study was compiled with the help of the Korean Climate Data Navigator (KCDN) (Sim, 2026; https://doi.org/10.5281/zenodo.23300923).

Links¶

• EN4 home page (Met Office Hadley Centre)
• EN.4.2.2 download page (with version history)
• EN.4.2.2 Product User Guide (PDF)
• Analysis file format
• Profile file format
• Analysis depth levels (depths.txt)
• EN4 FAQ
• EN4 terms and conditions
• EN.4.2.2 data sources