search

orcestra-campaign / weather

Weather briefings for the ORCESTRA field campaign.
5 stars 1 forks source link

Add ensemble forecasts #160

Closed HF7weatherman closed 2 months ago

HF7weatherman commented 2 months ago

This MR adds the spaghetti plot of the ICWV 48mm contour based on the IFS ensemble forecast, after @lkluft implemented its operational download. Furthermore, the spaghetti plots includes the daily climatology of the 48mm ICWV contour based on a 7-day running mean from ERA5 data (averaged over 2010 to 2022). Please see the example plot attached. I currently added the spaghetti plots on an additional slide in the weather briefing, which comes after the internal forecast slides. The structure of this slide is like the dust plots. As the annotation box had to be adjusted in the spaghetti plots, I took the chance to make the annotation box plot-dependent in general. Therefore, also the sfc_winds plot has now a more accurate annotation box, which addresses issue #147.

The additional infrastructure to load the ERA5 climatology is not very advanced yet and should potentially be put into a separate file instead of being located within icwv_ens.py. However, this refactoring should be addressed in a separate issue. Furthermore, the download of the climatological data is rather slow due to the large amount of data. Therefore, it would make sense to load both, the ERA5 data as well as the forecast data not within each plot, but once before the generation of all plots. This has the potential to save a lot of time during generating the briefing as it would avoid redundant loading of the same datasets. However, also this should be addressed in a separate issue. For now, please note that generating the briefing becomes longer by about 10 minutes due to the new spaghetti plots (FYI @karstenpeters, @sortega87, @Lucalino).

test_icwv_enfo

HF7weatherman commented 2 months ago

@karstenpeters @sortega87 @Lucalino technically, this branch is working. Thus, I'll merge it now, so we can include the new plots in tomorrow's weather briefing. However, as already written in the overview, there's still plenty of room for improvement and refactoring. In case you have suggestions for improvement, please let me know. I'll create new issues tomorrow for the shortcomings I already identified.