search

ajleeson / research_blog

Weekly updates of research progress
0 stars 0 forks source link

research_blog/2023/09/25/model-eval-old-ICs #41

Open utterances-bot opened 1 year ago

utterances-bot commented 1 year ago

Early look at model evaluation with old ICs | Weekly Research Update

This week I focused on three main tasks: Preliminary model evaluation Working on slides for EFM presentation Reading and processing Scully (2013) More details below. Model evaluation At the time of writing, the new model evlaluation run has made it into early August. So far, the new model evaluation run appears to have comparable or better skill versus the GRC run. Bottom DO predictions in South Sound and Main Basin seem to have improved significantly. However, the model is still overpredicting bottom DO in the far reaches of Hood Canal. In the coming week, I am interested in further exploring the spatial distribution of skill. Are there certain regions that the model is consistently performing worse in, and w

https://ajleeson.github.io/research_blog/2023/09/25/model-eval-old-ICs.html

parkermac commented 1 year ago

Nice work. To focus your evaluation more you could make versions of your property-property plots (like Figs. 4 and 5) for Hood Canal and Strait of Georgia.

Also it will be interesting to see how the rest of the year plays out, especially the late summer.

ahornerdevine commented 1 year ago

The comparison (and improvement) at PW and CI look really good. It seems reasonable that there is a different process that is important in south HC that is not being represented well. I agree that property-property plots in those sites will help.

Also, is this the official unveiling of your PhD color theme!?