-
Notifications
You must be signed in to change notification settings - Fork 4
Large ensembles
Summary: While decadal forecasting datasets aren't designed with UNSEEN analysis in mind, they tend to be larger (in terms of the number of times they simulate today +/- 10 years) than regular "large ensembles" of historical-into-future (e.g. 1850-2100) runs due to the multiple initial dates. CAFE is possibly the largest decadal forecasting ensemble available. The trade off is the relatively coarse resolution (so you're limited to certain, broad spatial extent extremes) as compared to high resolution efforts such as Weather@Home.
Deser et al (2020) is the latest overview paper on the history and future of large ensemble science.
Openly available "large" ensembles, approximately listed in order of how large they are:
https://portal.nersc.gov/c20c/
XX-member ensemble of the historical period with and without anthropogenic forcing.
https://www.cesm.ucar.edu/projects/community-projects/LENS/
40-member ensemble of fully-coupled CESM1 simulations for the period 1920-2100.
https://www.cesm.ucar.edu/projects/community-projects/LENS2/
100 members at 1 degree spatial resolution covering the period 1850-2100 under CMIP6 historical and SSP370 future radiative forcing scenarios.
https://www.cesm.ucar.edu/projects/community-projects/MMLEA/
The US CLIVAR Working Group on Large Ensembles has produced a centralized data archive of initial-condition Large Ensembles conducted with currently up to 8 CMIP5-class climate models. I think this collection is also referred to as "LE-MIP".
https://pcmdi.llnl.gov/CMIP6/ArchiveStatistics/esgf_data_holdings/CMIP/index.html
Depending on what timescale and variable you're interested in, there's up to 610 runs of the historical experiment available across the multi-model ensemble.
The MMLEA archive also includes includes a 1000 member “Observational Large Ensemble” (Obs-LE) for temperature (land only), precipitation (land only), and sea level pressure (global), which can be used to sample alternative but non-observed historical climate trajectories. These data are based on a statistical model and described in McKinnon and Deser (2018).
https://www.climateprediction.net/
2700 high resolution simulations of the Australia/NZ region for the years 2013-2015 (e.g. Black 2016).
https://www.cesm.ucar.edu/projects/community-projects/DPLE/
62 initialization times (November 1 of 1954, 1955, ..., 2014, 2015). For each start date, a 40-member ensemble was generated by randomly perturbing the atmospheric initial condition. The simulations were integrated forward for 122 months after initialization.
f6 has 96 ensemble members, 2 initial dates per year, 10 year lead time.
... there might be other datasets at the WMO Lead Centre for Annual-to-Decadal Climate Prediction
References:
- Black, M. T., & Karoly, D. J. (2016). Southern Australia’s Warmest October on Record: The Role of ENSO and Climate Change, Bulletin of the American Meteorological Society, 97(12), S118-S121. https://doi.org/10.1175/BAMS-D-16-0124.1
- Black, M. T., Karoly, D. J., Rosier, S. M., Dean, S. M., King, A. D., Massey, N. R., Sparrow, S. N., Bowery, A., Wallom, D., Jones, R. G., Otto, F. E. L., and Allen, M. R (2016). The weather@home regional climate modelling project for Australia and New Zealand, Geosci. Model Dev., 9, 3161–3176, https://doi.org/10.5194/gmd-9-3161-2016
- McKinnon, K. A., & Deser, C. (2018). Internal Variability and Regional Climate Trends in an Observational Large Ensemble, Journal of Climate, 31(17), 6783-6802. https://doi.org/10.1175/JCLI-D-17-0901.1
- Yeager, S. G., Danabasoglu, G., Rosenbloom, N. A., Strand, W., Bates, S. C., Meehl, G. A., Karspeck, A. R., Lindsay, K., Long, M. C., Teng, H., & Lovenduski, N. S. (2018). Predicting Near-Term Changes in the Earth System: A Large Ensemble of Initialized Decadal Prediction Simulations Using the Community Earth System Model, Bulletin of the American Meteorological Society, 99(9), 1867-1886. https://doi.org/10.1175/BAMS-D-17-0098.1