Strategy to Develop Modeling Requirements for Global Forecast Systems Based on Users’ Needs identified from User Feedback

The primary mission of the Analysis and Forecast Branch (a.k.a. AFS11) in the Analyze, Forecast, and Support Office of National Weather Service (NWS) is to address the field’s needs on Analysis and Forecasting by developing operational requirements for modeling systems used by the NWS forecasters, including global forecast systems such as GFS (Global Forecast System), GEFS (Global Ensemble Forecast System), CFS (Climate Forecast System) and also the forthcoming SFS (Seasonal Forecast System), which is being planned.

Developing requirements for global forecast systems is directly useful for meeting the needs of the Weather Enterprise by helping address and reduce systematic biases of the models. The requirements will also be indirectly useful for regional analysis and forecast systems, since the regional systems take the input (boundary and initial conditions) from global forecast systems. If, for example, a regional forecast system is requested to produce forecast information of good quality, despite the input condition from a global forecast system being of poor quality, the system is producing a right answer for a wrong reason. Moreover, some systematic biases of NWS’ forecast systems are commonly found both in regional and global forecast systems, which may be due to this adverse influence of global systems on regional systems. AFS11’s requirements are developed with this connection in mind and across systems of different scales so that biases are addressed in a coherent manner through linked requirements if possible.

Our approach to developing requirements is to first collect sample field cases, which illustrate deficiencies in the existing versions of GFS and GEFS, and to develop a user survey to collect feedback from both NOAA (National Oceanic and Atmospheric Administration) and non-NOAA users. Based on the case collection and evaluation, we developed and distributed an extensive user survey in early 2024 to collect feedback from various users throughout the Weather Enterprise (academia, private sector, retirees, as well as local and federal governments). Survey questions covered a broad range of topics, ranging from tropical cyclones to winter weather and warm season convection, as well as biases with basic fields such as temperature and dew point. We analyzed the survey data as a preparation to develop draft requirements for global forecast systems. The survey data show users’ needs both in short-term and extended-range forecasting.

This presentation will provide our strategy and an outline of this requirements development process. The results of the survey together with a draft form of requirements for global forecast systems will be presented in a companion paper (Hultquist et al.)