1. Overview of METexpress

1.1. The METplus suite

METexpress is an easy-to-use interface that displays plots of statistical verification metrics for the data that a user defines interactively. Prior to using METexpress, these verification metrics must generated from model output and “truth” data (usually observations or gridded model analyses) by the MET verification tools (https://met.readthedocs.io/en/latest/), and the output files produced by this MET verification must be loaded into a MET database. METexpress is not used to create this database, only view its contents. It allows a model developer to explore metrics about their model runs quickly and flexibly without relying on someone else producing pre-generated plots. The developer can slice and dice data in the way that best gives them insight into how their model performed.

METexpress is one component of the METplus verification tool suite, which is created and managed by the Developmental Testbed Center (DTC) and is the official verification tool for NOAA’s Unified Forecast System. Figure 1.1 shows a graphic of the entire METplus package and how METexpress integrates with the other tools.


Figure 1.1 Schematic diagram of the structure of METplus, including METexpress

In the simplest workflow for verifying model outputs, the model output data and “truth” or observation data are input into the MET software package which outputs statistical information in an ASCII format. That statistical information is then read into and stored in the METdatadb database using the METdbload tool. METexpress allows the user to specify various parameters for the plot they want to create and then queries the database to get the relevant information and creates a plot.

1.2. METexpress Features

METviewer performs a similar functionality of querying the database and producing a plot. However, METviewer has a lot more flexibility and many options that the user must know to specify, with the result that it requires more training to fully understand it and use it effectively and accurately. METexpress provides an alternative simplified, more intuitive interface, which guides the user through selection of parameters needed in order to produce a given type of plot. The expectation is that METviewer would be used by expert users who understand its extensive capabilities and advanced features, but that METexpress would provide an interface for users to quickly generate commonly used plots without having to gain the expertise needed to run METviewer. METexpress guides the user to define each parameter that is needed to produce a given type of plot, with default parameter values available also.

METexpress was developed at NOAA/OAR Global Systems Laboratory (GSL) based on a verification system developed in-house at GSL named the Model Analysis Tool Suite, or MATS. METexpress uses the basic framework of MATS but it has been modified to work with the METplus database.

It is very important to understand that METexpress can only produce plots based on the data that has been loaded into the METdatadb database.

The verification measures or statistics produced by METexpress follow definitions set by the MET package with input from the mathematical expertise within the Developmental Testbed Center (DTC). To learn more about the metrics and how to interpret them, please see the MET User Guide, particularly Appendix C about Verification Measures and additional material.

One of the features of METexpress is that, generally, it only presents the user with choices for data parameters that are valid for the data sets that the user has selected. For example, if the user selects the model “GFS”, the choices available for “variable” will be only the variables that have been loaded into the database for the GFS model. In this way, the user will not go through the process to select variables, heights, statistics, etc that don’t exist in the database and cannot be plotted. This is achieved in the interface by creating metadata (data that describes the data) in the database that describe the available parameters for each data set. The generation of metadata will be discussed later.