11/10/18 Model Verification

Nerd rating: Mild

There are a couple of ways I verify models before I start using them for a post. When it is stormy out I look at the latest model run and compare where the precipitation is compared to where the model’s simulated radar forecasted it to be. I also verify forecasted temps for a particular time. Since we have no local precipitation to track let’s look at how horribly the models verified overnight.

First current 6 am temps, this shows actual temps from reporting stations in Colorado, as you can see at DRO  it is 13 degrees.

11102zrtma_t2m_colorado

Now a look at what the models forecasted, first the GFS. 25 degrees! Horrible forecast.

gfs_t2m_b_colorado_2

Euro 22 degrees, it still missed by 9.

ecmwf_t2min_colorado_3

Now the HiRes model which in my opinion has one accurate run a day and that is the one due out in an hour. I usually never use the overnight run but here it is. 25 degrees! Huge miss.

hires_t2m_colorado_7.png

What does this mean? It means the models had a bad night. When you read in one of my posts that a model didn’t verify well so I have to wait for another run of the model, this is what I am talking about. Bottom line, I don’t trust those models for data for the very short term (6-12 hours). It does not have a dramatic effect on the longer-term outlook.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s