Data for Uncertainty Quantification of a Machine Learning Subgrid-Scale Parameterization for Atmospheric Gravity Waves
Abstract/Contents
- Abstract
Contains data generated in Mansfield & Sheshadri (2024): Uncertainty Quantification of a Machine Learning Subgrid-Scale Parameterization for Atmospheric Gravity Waves, JAMES.
Data includes climate model integrations from MiMA (Model of an idealized Moist Atmosphere) using the spectral gravity wave parameterization from Alexander & Dunkerton (1999; AD99) used for training, validation and testing, all neural network torchscript models, and MiMA files required to initialize the online simulations. We also include the post-processed zonal mean winds and gravity wave drag in the QBO and polar vortex regions for all neural network simulations and the first three years of offline and online simulations for all ensemble members.
Directory structure:
*train_wavenet/ contains training, validation and test data from AD99 MiMA integrations described above. Outputs are saved as netcdf4 files.
*mima_files/ contains all files required to initialize the NN MiMA simulations, including compiled MiMA using FTorch, namelist, diagnostic tables, input files, and restart files. To run MiMA, all these files must be copied into the run directory for each ensemble member.
* wavenet_1_seed{100-129}/ contains the first three years of online MiMA simulations each NN ensemble member, and post-processed zonal mean winds and gravity wave drag for all pressure levels in the QBO region (5 deg S to 5 deg N), and polar vortex regions (60 deg N for Northern hemisphere and 60 deg S for Southern hemisphere). Within this directory,
*MODELS/ contains the torchscript files in the correct format to be read by FTorch in MiMA.
* zonal_offline/ contains the first three years of offline zonal neural network tests.
* meridional_offline/ contains the first three years of offline meridional neural network tests.
Description
Type of resource | Dataset, three dimensional object, cartographic |
---|---|
Publication date | May 16, 2024 |
Creators/Contributors
Author | Mansfield, Laura |
![]() |
---|---|---|
Author | Sheshadri, Aditi |
Subjects
Subject | Machine learning |
---|---|
Subject | Gravity waves |
Subject | Atmospheric circulation > Models |
Subject | Stratosphere > Simulation methods |
Subject | Polar vortex |
Subject | Atmospheric circulation |
Genre | Data |
Genre | 3d model |
Genre | Geospatial data |
Genre | Data sets |
Genre | Dataset |
Genre | Three-dimensional scan |
Genre | Cartographic dataset |
Bibliographic information
Related item |
|
---|---|
DOI | https://doi.org/10.25740/zv875tm6846, https://doi.org/10.25740/zv875tm6846 |
Location | https://purl.stanford.edu/zv875tm6846 |
Access conditions
- Use and reproduction
- User agrees that, where applicable, content will not be used to identify or to otherwise infringe the privacy or confidentiality rights of individuals. Content distributed via the Stanford Digital Repository may be subject to additional license and use restrictions applied by the depositor.
- License
- This work is licensed under a Creative Commons Attribution 4.0 International license (CC BY).
Preferred citation
- Preferred citation
- Mansfield, L. and Sheshadri, A. (2024). Data for Uncertainty Quantification of a Machine Learning Subgrid-Scale Parameterization for Atmospheric Gravity Waves. Stanford Digital Repository. Available at https://purl.stanford.edu/zv875tm6846. https://doi.org/10.25740/zv875tm6846.
Collection
Stanford Research Data
View other items in this collection in SearchWorksContact information
Also listed in
Loading usage metrics...