The Model Evaluation Tools (MET) were developed and released to the community by the Developmental Testbed Center more than a decade ago. At that time, the tools included computation of traditional statistics for continuous fields (temperature, pressure, and height) and dichotomous fields (precipitation, clouds, fog, and high impact events). MET also included a spatial verification method, using objects, which complemented traditional statistics.
Over the past five years, new flexibility and diagnostic options expanded MET capabilities. Support for calling a Python script from within MET was added to give researchers more control over the fields supported by MET and to allow for exploration of new methods. Additionally, a suite of Python wrappers now provides low-level automation for verification tasks. The resulting enhanced MET system is called METplus and is being adopted by DTC partners that want to tap into the power of an extensive toolset and unified verification capability.
Upcoming enhancements are driven by many projects within the DTC and through community contributions:
- More process-oriented diagnostics ranging from microphysical scales to sub-seasonal to seasonal scales.
- A renewed focus on enhancing MET-TC (Tropical Cyclone) to provide diagnostics in “cyclone-space.”
- Additional capability for: evaluating high-impact weather and atmospheric composition events, the use of satellite data, and evaluating components of a coupled prediction system.
Look for many of these enhancements in the METplus v3.0 release in late fall 2019.
