search

NOAA-GSL / VxIngest

Other
2 stars 0 forks source link

observation-space verification of the DA ensemble performance #41

Open bonnystrong opened 3 years ago

bonnystrong commented 3 years ago

Curtis: Thanks for asking and actually we do run an observation-space verification of the DA ensemble performance. We use the GSI diagnostic output to feed the verification for all ensemble members to include the spread, error and bias information. Here’s an example of what that looks like for a METARs. Technically this is the forecast ensemble, but we can execute the same thing for the DA ensemble (just 1-hr forecasts for more members). We’d definitely like to have this verification capability as well.

Google doc: https://docs.google.com/document/d/1irOgr_Vqo7k28v_5ufInITEj-Va2aMZ0TOhRSj3nWmg/edit?usp=sharing

Screen Shot 2021-07-13 at 3 53 34 PM

bonnystrong commented 1 year ago

Requires info about observation error.

Jeff: METplus has a first effort at this with something on obs error from a static table (which is included in METplus). Have not looked at METAR.

Curtis: action now is to validate capability. Raise priority for just the first step to investigate what's available and get an estimate of time to implement.

bonnystrong commented 1 year ago

Jeff: need to review if obs error in MET. That is the High priority part of the bigger task.

github-actions[bot] commented 1 year ago

This issue is stale because it has been open 90 days with no activity.