Introduction

This dataset is a transient run of WACCM-X between 1950 and 2024 inclusive, with hourly output particularly focused on thermospheric and ionospheric variables, designed to provide model data on climatological timescales as well as shorter term events such as geomagnetic storms. WACCM-X is a whole atmosphere numerical model, calculating the physics, chemistry and dynamics of the atmosphere from ground-level to the upper atmosphere (around 500 km) and allows ground-level effects, such as carbon dioxide emissions, to propagate up through the atmosphere, providing a powerful tool for investigating the climatology of the upper atmosphere. Applications of the dataset include event and climatological studies, validation against observations and other models, machine learning, driving of other models' boundary layers, and more.

Dataset

Files are in netcdf (HDF5) format. They are split into daily output, named WACCMXhistory.cam.h1.YYYY-MM-DD-00000.nc, where YYYY-MM-DD is replaced with the date the file corresponds to. These files are split into yearly folders, from 1950 through to 2024 inclusive.

Further details of the data format and a form for download are available from the Data Download page.

Model Information

The Whole Atmosphere Community Climate Model with thermosphere and ionosphere extension (WACCM-X) version 2.1 as part of the Community Earth System Model (CESM) version 2.1.3.

Liu, Han-Li, et al. "Development and validation of the Whole Atmosphere Community Climate Model with thermosphere and ionosphere extension (WACCM-X 2.0)." Journal of Advances in Modeling Earth Systems 10.2 (2018): 381-402. https://doi.org/10.1002/2017MS001232

• https://www2.hao.ucar.edu/modeling/waccm-x
• https://www.cesm.ucar.edu/

Model Setup

All driving files were left as the default, except those listed in this section, with justification given.

WACCM-X's geomagnetic field is driven by default by the 12th release of International Geomagnetic Reference Field (IGRF-12). The newer IGRF-13 has since been released, so the coefficients were updated. This only affect years after 2015 as the 2015-2020 main field reflects observed values and the 2020-2025 secular variation is updated.

The solar and geomagnetic indices used to drive the model are available from https://serene.bham.ac.uk/resources/Indices/ . The F10.7 solar activity index was cleaned to remove solar flares using a Hampel identifier with a sliding window of 15 days and outliers removed if they are larger than threes multiples of the median absolute deviation.

All other changes to driving files were made due to the default not starting early enough. Greenhouse gas emissions, sulphates, solar irradiance, and prescribed aerosol were changed to files used by the Chemistry-Climate Model Initiative (CCMI) which went back to 1950. During the spin-up year of 1949, any driving files missing this year used data from the year 1950. This includes the CCMI ones above, galactic cosmic ray ionization parameterization, and the prescribed stratospheric aerosol.

Files covering ion-pair production rates only went back to 1963, as observations from NOAA G11 only start from this year. In order to fill the 1950-1963 data gap, rates from solar cycle 21 (1976-03-01 to 1986-09-01) were repeated for earlier solar cycles.

Due to computing constraints, the simulation was split into 3, starting from 1949, 1973 and 1997 with one year of spin-up each before storing the data contained in this repository. The initial files for each of these used the default 2000 run, but with all CO₂ and CO values scaled by the ratio of the average ground-level CO₂ concentration of the starting year against the year 2000.

Further details are provided in the paper entitled “ WACCM-X simulation of 1950 through 2024 at an hourly resolution”.

Funding and acknowledgments

This dataset was created on UKRI's ARCHER2 HPC (https://www.discover.ukri.org/archer2/index.html) and updated to recent years on the University of Birmingham's BlueBEAR HPC (https://www.birmingham.ac.uk/research/arc/bear/bluebear). It was completed with compute time obtained by Oliver Allanson (NE/V013963/1 and NE/V013963/2) under the rad-sat project (NE/P017274/1 and NE/P017274/2). The model was run by Matthew Brown under the SWIMMR-T project (NE/V002643/1 and NE/V002708/1) to investigate thermosphere neutral density climatology for space debris applications, with support from Sean Elvidge and David Themens (NE/W003368/1 and NE/W003147/1).

Acknowledgements

This work used the ARCHER2 UK National Supercomputing Service (https://www.archer2.ac.uk) and the University of Birmingham's BlueBEAR HPC service (https://www.birmingham.ac.uk/bear). M. K. Brown and S. Elvidge are supported by the UK Space Weather Instrumentation, Measurement, Modelling and Risk (SWIMMR) Programme, Natural Environment Research Council (NERC) Grants NE/V002643/1 and NE/V002708/1.

O. Allanson would like to acknowledge the United Kingdom Research and Innovation (UKRI) NERC Independent Research Fellowship NE/V013963/1 and NE/V013963/2; and the NERC Highlight Topic Grant NE/P017274/1 and NE/P017274/2 (Rad-Sat).

D.R. Themens would like to acknowledge the support of the UK NERC DRIIVE and FINESSE EISCAT3D Highlight Topic Grants NE/W003368/1 and NE/W003147/1.