Skip to content

JRA-3Q

Japanese Reanalysis for Three Quarters of a Century (JRA-3Q)

Available

Get notice alertsCompare with other datasetsReport an error

Current status

Available and current; JRA-3Q is the official successor to JRA-55. At NCAR GDEX the main dataset grows monthly and the near-real-time dataset every few days. DIAS in Japan has only partly resumed service after its long-term maintenance in 2026, and currently offers only data dated 24 June 2026 or later. The licence is non-commercial (CC BY-NC-SA 4.0); for commercial use, obtain the data through the Japan Meteorological Business Support Center (JMBSC).

At a glance

Item Details
Section Atmosphere › Reanalyses
Type Reanalysis
Domains Atmosphere, Land
Provider Japan Meteorological Agency (JMA)
Period 1947-09 ~ present
Coverage Global
Horizontal resolution Model resolution TL479 (about 40 km) / 1.25° × 1.25° latitude–longitude grid (288 × 145) / 0.375° × 0.375° grid based on the model grid (960 × 480)
Vertical resolution 100 model levels (top at 0.01 hPa) / 45 isobaric surfaces (1000–0.01 hPa) / 36 isentropic surfaces (270–4,000 K)
Temporal resolution Analyses every 6 hours, surface diagnostics and fluxes hourly (monthly and monthly diurnal statistics provided separately)
Update frequency Monthly (NCAR main dataset) / near-real-time dataset updated continually with a lag of a few days
Licence Creative Commons Attribution Non Commercial Share Alike 4.0 International (CC BY-NC-SA 4.0) — for commercial use, purchase the data through JMBSC
Account required Yes

Note on period: From 1 September 1947. The NCAR main dataset (d640000) is extended month by month (through September 2026 when checked in October 2026); the most recent days to weeks appear first in the near-real-time dataset (d640001). The sub-product JRA-3Q-COBE covers only June 1985 to December 1990.

How to read the resolution
  • Horizontal resolution lists both the native model grid and the grid you download, as given in the official documentation. The same dataset can come on different grids from different providers or products, so check the grid of your downloaded file.
  • Grid spacing is not the effective resolution. A provided grid that is finer than the native one carries no extra information.
  • Intervals and whether a value is instantaneous, mean or accumulated differ between variables — see 'Available variables' below.
  • The 'provided resolution' in lists and the comparison table is a simplified km/degree value for quick comparison. Use this page for exact values.

Overview

JRA-3Q is JMA's third long-term global atmospheric reanalysis, covering September 1947 to the present. It uses a data assimilation system based on JMA's operational set-up as of December 2018 (TL479, 4D-Var) and assimilates newly rescued historical observations and reprocessed satellite data, alleviating many of the deficiencies of JRA-55, such as the energy imbalance at the top of the atmosphere and the surface and the artificial weakening trend in tropical cyclone intensity. SST comes from COBE-SST2 until May 1985 and from MGDSST from June 1985; JRA-3Q-COBE was released alongside so the effect of that switch can be assessed.

What changed from the previous dataset

Available variables

Main variables from the JMA document "JRA-3Q Product Format (1.25-degree latitude/longitude grid)" and the NCAR GDEX (d640000) NetCDF files. Names are the JMA short names that begin the GDEX file and variable names (e.g. tmp2m-hgt-an-ll125 → tmp2m). The originals are GRIB2; GDEX converts them to NetCDF4, one variable per file. Note that precipitation and evaporation units differ from JRA-55 (mm/day).

Time-stamp convention — instantaneous, mean or accumulated? — click to show

Analyses (anl_*) are instantaneous values every 6 hours (00, 06, 12, 18 UTC), and forecast diagnostics are hourly. fcst_surf125 and bnd_ocean125 hold hourly instantaneous values; fcst_phy2m125 (precipitation, fluxes, radiation) holds 1-hour averages. Precipitation and evaporation are 1-hour mean rates (kg m-2 s-1), not accumulations.

  • In the NCAR GDEX files the variable attribute time_step states the temporal nature, e.g. "6-hour" (analyses), "1-hour instantaneous" (instantaneous forecasts), "1-hour average" (averaged diagnostics) and "24-hour" (snow analysis). Check this attribute when in doubt.
  • The snow depth analysis (anl_snow125) is daily, at 18 UTC.
  • The JMA format document does not state how the 1-hour averages are time-stamped. The GDEX daily files are labelled 00–23 UTC, which suggests that, as in JRA-55, the time stamp marks the start of the averaging interval (our inference). Check this yourself if precise timing matters.
  • Which forecast window the averaged diagnostics come from is only mentioned in the format document's description of climatological normals ("averages from the beginning of forecasts up to six hours for average diagnostic fields").
  • In minmax_surf (Gaussian grid), 2 m maximum and minimum temperature and maximum wind speed are maxima and minima over 1-hour intervals (GDEX time_step "1-hour maximum").
  • To compare with the JRA-55 3-hour means (mm/day), average three 1-hour means and multiply by 86400.

Source: JRA-3Q 1.25° format document + NCAR GDEX d640000 file metadata (time_step)

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

Isobaric analysis (anl_p125, 6-hourly)

Variable Short name Units Time convention Notes
Geopotential height hgt gpm instantaneous (every 6 hours)
Temperature tmp K instantaneous (every 6 hours)
u-component of wind ugrd m s-1 instantaneous (every 6 hours)
v-component of wind vgrd m s-1 instantaneous (every 6 hours)
Vertical velocity (pressure) vvel Pa s-1 instantaneous (every 6 hours)
Specific humidity spfh kg kg-1 instantaneous (every 6 hours)
Relative humidity rh % instantaneous (every 6 hours)
Dewpoint depression (or deficit) depr K instantaneous (every 6 hours)
Relative vorticity relv s-1 instantaneous (every 6 hours)
Stream function · velocity potential strm, vpot m2 s-1 instantaneous (every 6 hours)

Surface analysis (anl_surf125, 6-hourly)

Variable Short name Units Time convention Notes
Temperature at 2 m tmp2m K instantaneous (every 6 hours)
Specific humidity at 2 m spfh2m kg kg-1 instantaneous (every 6 hours)
Relative humidity at 2 m rh2m % instantaneous (every 6 hours)
u- and v-components of wind at 10 m ugrd10m, vgrd10m m s-1 instantaneous (every 6 hours)
Pressure (ground or water surface) pres Pa instantaneous (every 6 hours)
Pressure reduced to MSL prmsl Pa instantaneous (every 6 hours)
Total column integrated water vapour tciwv kg m-2 instantaneous (every 6 hours)
Water equivalent of accumulated snow depth weasd kg m-2 instantaneous (every 6 hours)

Fluxes and diagnostics — 1-hour average forecasts (fcst_phy2m125)

Variable Short name Units Time convention Notes
Total precipitation rate tprate kg m-2 s-1 1-hour mean GDEX time_step "1-hour average" confirmed. The units are a rate, not an accumulation
Convective · large scale precipitation rate cprat, lsprate kg m-2 s-1 1-hour mean
Total snowfall rate, water equivalent tsrwe kg m-2 s-1 1-hour mean
Evaporation rate evarate kg m-2 s-1 1-hour mean
Latent · sensible heat net flux lhtfl, shtfl W m-2 1-hour mean
Momentum flux, u- and v-components uflx, vflx N m-2 1-hour mean
Downward short-wave · long-wave radiation flux dswrf, dlwrf W m-2 1-hour mean
Upward short-wave · long-wave radiation flux (surface and top of atmosphere) uswrf, ulwrf W m-2 1-hour mean
Vertically integrated zonal · meridional water vapour flux tuwv, tvwv kg m-1 s-1 1-hour mean

Clouds, ozone, snow — 1-hour instantaneous forecasts (fcst_surf125)

Variable Short name Units Time convention Notes
Total · low · medium · high cloud cover tcdc, lcdc, mcdc, hcdc % instantaneous (every hour)
Total ozone tozne DU instantaneous (every hour)
Snow depth snod m instantaneous (every hour)

Ocean boundary conditions (bnd_ocean125)

Variable Short name Units Time convention Notes
Water temperature wtmp K instantaneous (every hour)
Ice cover icec proportion instantaneous (every hour)

The full variable list is in the official documentation.

How to get the data

Practical tips

  • Precipitation and evaporation are in kg m-2 s-1 (= mm s-1), not mm day-1 as in JRA-55. Multiply by 86400 to convert to mm/day before comparing with JRA-55.
  • The SST input switches from COBE-SST2 (1°) to MGDSST (0.25°) in June 1985. If your trend or variability analysis spans that date, use JRA-3Q-COBE to check its effect.
  • JMA has reported a local quality issue from June 1985 to April 2013 affecting ice cover, surface fluxes and 2 m temperature within about 250 km of the sea-ice edge. Check it if you study the polar sea-ice margin.
  • The metadata inside the GDEX files give the pages of Kosaka et al. (2024) as 102, 1-168, but the official J-STAGE record is 102(1), 49–109.

Notices and change history

2026-08-19 · Warning · DIAS Notice of Partial Resumption — JRA-3Q only data dated 24 June 2026 or later

DIAS, which had been under long-term maintenance, resumed downloads of part of the JRA-3Q data: "only data dated 24 June 2026 or later". Republication dates for the remaining data have not yet been determined. Use NCAR GDEX for earlier periods.

Read the original

2026-04-28 · Warning · DIAS Service Outage — all datasets inaccessible

DIAS announced that it is in long-term maintenance, that all systems except the website are unavailable, and that "the expected date of service restoration has not yet been determined." The JMA JRA-3Q page likewise says DIAS datasets are inaccessible.

Read the original

2025-12-02 · Info · JMA reanalysis website moved

The JRA project website moved from jra.kishou.go.jp to www.data.jma.go.jp/jra/html/. The old address redirects automatically for the time being; update your bookmarks.

Read the original

2023-12-05 · Info · Near real-time JRA-3Q data available from NCAR

Near real-time JRA-3Q data became available from NCAR. Around the same time JMA announced that near real-time JRA-55 updates would end at the end of January 2024.

Read the original

2023-08-10 · Info · Full period and JRA-3Q-COBE available from DIAS

Data for September 1947 to December 1990 and from May 2013 onward were added to DIAS, completing the record, and the sub-product JRA-3Q-COBE was released at the same time.

Read the original

2022-12-09 · Warning · Local issues with sea ice parameters

From June 1985 to April 2013, sea-ice areas were expanded by several grid cells before conversion to the model grid, so ice cover that should be 0.0 is reported as an average of 0.15 within around 250 km of the ice edge. Surface radiation and turbulent heat fluxes and near-surface temperature and humidity nearby are also slightly affected. JMA describes the effects as local and minor, a small fraction of interannual variability.

Read the original

References for your paper

Double-check before using in a paper

The citations below were compiled by the maintainer and may contain errors. Always double-check DOIs, authors, years and access-date formats against the provider's official dataset page and your journal's guidelines.

Dataset citation

NCAR GDEX main dataset (d640000) — DOI 10.5065/AVTZ-1H78

Japan Meteorological Agency (2023): Japanese Reanalysis for Three Quarters of a Century (JRA-3Q). Research Data Archive at the National Center for Atmospheric Research, Computational and Information Systems Laboratory. https://doi.org/10.5065/AVTZ-1H78 (Accessed DD MMM YYYY)

Show BibTeX 
@misc{jra3q_d640000,
  author    = {{Japan Meteorological Agency}},
  title     = {Japanese Reanalysis for Three Quarters of a Century ({JRA-3Q})},
  year      = {2023},
  publisher = {Research Data Archive at the National Center for Atmospheric Research, Computational and Information Systems Laboratory},
  doi       = {10.5065/AVTZ-1H78},
  note      = {Accessed DD MMM YYYY}
}

NCAR GDEX near-real-time dataset (d640001) — DOI 10.5065/E3QQ-0F24

Japan Meteorological Agency (2023): Near Real-Time Japanese Reanalysis for Three Quarters of a Century (JRA-3Q). Research Data Archive at the National Center for Atmospheric Research, Computational and Information Systems Laboratory. https://doi.org/10.5065/E3QQ-0F24 (Accessed DD MMM YYYY)

NCAR GDEX JRA-3Q-COBE (d640006) — DOI 10.5065/RM7V-3X70

Japan Meteorological Agency (2024): Japanese Reanalysis for Three Quarters of a Century - COBE (JRA-3Q-COBE). Research Data Archive at the National Center for Atmospheric Research, Computational and Information Systems Laboratory. https://doi.org/10.5065/RM7V-3X70 (Accessed DD MMM YYYY)

Reference papers

JRA-3Q comprehensive report (the paper JMA asks users to cite) — DOI 10.2151/jmsj.2024-004

Kosaka, Y., Kobayashi, S., Harada, Y., Kobayashi, C., Naoe, H., Yoshimoto, K., Harada, M., Goto, N., Chiba, J., Miyaoka, K., Sekiguchi, R., Deushi, M., Kamahori, H., Nakaegawa, T., Tanaka, T. Y., Tokuhiro, T., Sato, Y., Matsushita, Y., Onogi, K. (2024): The JRA-3Q Reanalysis. J. Meteor. Soc. Japan, 102(1), 49–109.

Show BibTeX 
@article{kosaka2024jra3q,
  author  = {Kosaka, Yu and Kobayashi, Shinya and Harada, Yayoi and Kobayashi, Chiaki and Naoe, Hiroaki and Yoshimoto, Koichi and Harada, Masashi and Goto, Naochika and Chiba, Jotaro and Miyaoka, Kengo and Sekiguchi, Ryohei and Deushi, Makoto and Kamahori, Hirotaka and Nakaegawa, Toshiyuki and Tanaka, Taichu Y. and Tokuhiro, Takayuki and Sato, Yoshiaki and Matsushita, Yasuhiro and Onogi, Kazutoshi},
  title   = {The {JRA-3Q} Reanalysis},
  journal = {Journal of the Meteorological Society of Japan. Ser. II},
  volume  = {102},
  number  = {1},
  pages   = {49--109},
  year    = {2024},
  doi     = {10.2151/jmsj.2024-004}
}

Acknowledgements

JMA states that references to JRA-3Q data in scientific or technical papers should be cited to the JRA-3Q comprehensive report (Kosaka et al., 2024), and that JRA-3Q data should be acknowledged, for example: "This report references JRA-3Q reanalysis data from the Japan Meteorological Agency." If you obtained the data from NCAR or DIAS, also give the dataset citation shown on that organization's metadata page (for DIAS, doi:10.20783/DIAS.645).

If this page helped, please acknowledge the Korean Climate Data Navigator

Please add the sentence below to the Acknowledgements of your paper.

Acknowledgements

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).