Merge pull request #1094 from CliMA/kp/pressure-postprocessing

Add postprocessing for 4D variables (lat, lon, time, pressure level)

NEWS.md

@@ -21,6 +21,12 @@ by identifying where elevation is greater than 0. Note, this can lead to
 misidentification of ocean in some areas of the globe that are inland but below
 sea level (Dead Sea, Death Valley, ...).
 
+### Leaderboard for variables over longitude, latitude, time, and pressure - PR [#1094](https://github.com/CliMA/ClimaCoupler.jl/pull/1094)
+
+As a part of the post processing pipeline, bias plots for variables at the
+pressure levels of 850.0, 500.0, 250.0 hPa and bias plots over latitude and
+pressure levels are being created.
+
 
 
 ### Code cleanup

docs/src/leaderboard.md

@@ -9,10 +9,13 @@ preprocessing variables of interest are done in `data_sources.jl` and computing
 and plotting are done in `leaderboard.jl`. To add a new variable, you ideally only need to
 modify `data_sources.jl`.
 
-#### Add a new variable to the bias plots
-If you want to add a new variable to the bias plots, you add the variable to `sim_var_dict`,
-`obs_var_dict`, `compare_vars_biases_groups`, and optionally
-`compare_vars_biases_plot_extrema`.
+#### Add a new 3D variable to the bias plots
+If you want to add a new 3D variable defined over latitude, longitude, and time to the bias
+plots, you add the variable to `sim_var_dict`, `obs_var_dict`, `compare_vars_biases_groups`,
+and optionally `compare_vars_biases_plot_extrema`. The variables `sim_var_dict`,
+`obs_var_dict`, `compare_vars_biases_groups`, `compare_vars_biases_plot_extrema` are in the
+function `get_sim_var_dict`, `get_obs_var_dict`, `get_compare_vars_biases_groups`, and
+`get_compare_vars_biases_plot_extrema` respectively.
 
 The dictionaries `sim_var_dict` and `obs_var_dict` map short names to an anonymous function
 that returns a [`OutputVar`](https://clima.github.io/ClimaAnalysis.jl/dev/var/). Both
@@ -48,3 +51,14 @@ must be initialized for each variable of interest. The CliMA model is added with
 the `RMSEVariable`. It is assumed that the `RMSEVariable` contains only the columns "DJF",
 "MAM", "JJA", "SON", and "ANN" in that order. The file `leaderboard.jl` will load the
 appropriate data into the `RMSEVariable`.
+
+### Add a new variable to compare against observations in pressure coordinates
+To add a new variable, you only need to modify the variable `sim_var_pfull_dict` in the
+function `get_sim_var_in_pfull_dict`, the variable `obs_var_dict` in the function
+`get_obs_var_in_pfull_dict`, and the variable `compare_vars_biases_plot_extrema` in the
+function `get_compare_vars_biases_plot_extrema_pfull`. The variables and functions are
+defined exactly the same as their analogous versions in the section above.
+
+It is expected that the dimensions of the variables are time, latitude, longitude, and
+pressure in no particular order and the units for the pressure dimension is expected to be
+`hPa`.

experiments/ClimaEarth/Artifacts.toml

@@ -58,3 +58,6 @@ git-tree-sha1 = "cbbc6b3752d9cb9b667ec33cfbeb46819f8db418"
     [[landsea_mask_1deg.download]]
     sha256 = "3722b553c2fdf28a6574aea2e0b167d16ab050f34e5969ada45625b3a3ecb6da"
     url = "https://caltech.box.com/shared/static/b3u4dv0dsoswvqgp8y63bzj7awbhwztd.gz"
+
+[era5_monthly_averages_pressure_levels_1979_2024]
+git-tree-sha1 = "dd7ee9500805f590f8fc4c00bfc373eeb7a01587"