602 Evaluation of the Grell-Freitas Convective Scheme within the NOAA Environmental Modeling System (NEMS)-based Global Spectral Model (GSM)

Tuesday, 24 January 2017
4E (Washington State Convention Center )
Michelle Harrold, NCAR, Boulder, CO; and J. K. Henderson, J. K. Wolff, C. Holt, L. R. Bernardet, H. Jiang, and L. B. Nance

The Global Model Test Bed (GMTB) is a newly established group within the Developmental Testbed Center (DTC) with the primary goal of supporting transition of research to operations for global numerical weather prediction (NWP). One focus area of GMTB is to provide tools and infrastructure for an automated end-to-end workflow system, which includes running a global model, post-processor, verification, and generation of diagnostics. Current workflow capabilities allow for running the Global Spectral Model (GSM); however, with the implementation of the Finite-Volume Cubed-Sphere (FV3) Dynamical Core, the GMTB will transition to adopting the new dynamical core. To showcase the capabilities the GMTB has established, as well as to generate information that aides in evidence-based decision making, a test was conducted to compare the Global Forecast System’s (GFS) operational convective scheme [Simplified Arakawa-Schubert (SAS)] to an experimental configuration using the scale-aware Grell-Freitas (GF) convective scheme. The development of scale-aware convective parameterizations is an active area of research in both the operational and research NWP communities, making the results from this test highly relevant.
Upon successful implementation of the GF scheme in the GSM, both configurations were run over an identical set of cases, spanning both a warm season (June, July, August 2015) and cool season (December 2015, January, February 2016). Using initial conditions from GFS reforecasts, the GSM was run at a resolution of T574, and forecasts were launched once a day at 00 UTC, producing output every 6 hours. This presentation will focus on the statistical assessment of the forecast performance of the two configurations for each season. Objective verification was computed for standard surface and upper air variables for a number of metrics and stratified by forecast lead time, vertical level, regional area, and season. Examples of time-averaged diagnostics (e.g., seasonal precipitation and diabatic heating rate) will also be presented. In addition, a companion paper (Jiang et al.) will be presented to highlight a case study from this extended test.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner