January 24, 2022 1:16 pm

METplus verification software for diagnosing model biases

Francisco Martin Leon 2 min
COMET module cover commented

Using METplus Verification Software to Diagnose Model Biases

About this module

You are working on an analysis of the output data for last year’s model, which has been very active. You seem to have detected a trend in the data, but you are having a hard time confirming your theories about it. As the delivery date approaches, the tension mounts. You have large amounts of data stored yet to be analyzed. To finalize the results of your study, you must determine if the prognoses have improved and diagnose the reason.

This lesson will help you understand how to use the verification software to evaluate the output of the model against a set of verification data. We will use METplus to show you how to analyze forecasts for a specific location as well as forecasts focused on a feature.

English language
TOPICS Numerical modeling (PNT)
STUDY TIME .50 – .75 h
PUBLICATION DATE January 3, 2022


  • List the advantages METplus diagnostic tools offer over others, such as Python, for verifying feature-centric forecasts.
  • Interpret the statistical output of the basic state variables generated by METplus.
    – Apply the interpretation of METplus statistical outputs to a new case study.
  • Evaluate the feature-centered output of METplus by comparing the model output with a set of verification data.
    – Identify discrepancies.

Go to the module, here.



Comments (0)

Leave a Reply

Your email address will not be published. Required fields are marked *