cmip6 dashboard

[phaustin]

No description provided.

[phaustin]

• Climate education grant
  https://drive.google.com/file/d/1cZNnhXVfhpqACpKvj26sDRE2PFCsqmK8/view

• Archer cmip5 climate maps
  http://climatemodels.uchicago.edu/maps/

• tools:
  http://xarray.pydata.org/en/stable/
  https://github.com/mwouts/jupytext
  https://zarr.readthedocs.io/en/latest/?badge=latest

• cmip6 information

  • google dataset doc
  • cmip6 explainer

[phaustin]

• write an introductory notebook to subset and plot cmip 6 model fields on a map
• add widgets to reproduce the behavior of http://climatemodels.uchicago.edu/maps/
• turn the notebook into a plotly dashboard
• get familiar with xarray by 1) installing the pythia datasets: https://anaconda.org/conda-forge/pythia-datasets and 2) doing http://www.atmos.albany.edu/facstaff/ktyle/atm533/core/week7/01_Xarray_Intro.html

[phaustin]

Some geoinformtion links:

basic coordinate reference systems:
https://a301_web.eoas.ubc.ca/week4/cartopy_mapping_h5.html#week4-coords

cartopy utm example:
https://a301_web.eoas.ubc.ca/week10/demo_cartopy_extent.html

satellite raster map:
https://a301_web.eoas.ubc.ca/week9/vancouver_visible.html

plotly radar raster
https://plotly.com/python/mapbox-layers/

cartopy plotly example:
https://stackoverflow.com/questions/58392381/using-cartopy-instead-of-basemap-to-create-map-traces-in-dash-app

vector shape files
https://a301_web.eoas.ubc.ca/week11/subset_map.html

affine transform example:
https://a301_web.eoas.ubc.ca/week11/assign6b_sol.html

mapbox tile crs:
https://epsg.io/3857

[JacobMcFarlane]

https://github.com/JacobMcFarlane/cmpi_6_dash.git

[phaustin]

For April 12: 1) phil will put plotly dash server on node07 2) jacob will write a method to get mean and variance from a field for a selected region

[phaustin]

Next steps:

• display histogram of field for selected region
• marginal distributions with something like https://seaborn.pydata.org/examples/hexbin_marginals.html
• compare histograms from two different climate runs for same region

[phaustin]

initial models and scenarios:

canesm2 (canada)
cesm2 (ncar)
hadgem3 (uk)

scenarios:
picontrol
historical
ssp245
ssp285

[phaustin]

1. pick a case with 1) regional corners 2) model runs 3) time period 4) scenario 5) model values --??? capture in a yaml, json, toml disk file
2. write a script that reads the case, goes to the cloud and downloads model files and saves as zarr or netcdf
3. update the dashboard so it can read the case file, and load local data using info in the case file
4. sample case: bc, surface air temperature, historical, 10-40 ensemble members

[phaustin]

New todo list for Jamie

• Xarray -- install xarray and https://anaconda.org/conda-forge/pythia-datasets
• do the tutorial: http://www.atmos.albany.edu/facstaff/ktyle/atm533/core/week7/01_Xarray_Intro.html
• do the 4 part pandas tutorial: https://github.com/phaustin/pandas_yvr along with http://www.atmos.albany.edu/facstaff/ktyle/atm533/core/pandas/pandas.html
• do the intake demo: http://gallery.pangeo.io/repos/NCAR/cesm-lens-aws/notebooks/EnhancedIntakeCatalogDemo.html
• refactor the groupby shown in https://github.com/phaustin/cmip6_book/blob/precip_plots/precip.py#L107-L124 so that we can get an anual precip cycle for a decade/region
• use this function(s) to compare annual cycle for some scenarios and models in BC, tropics and subtropics