93 Assimilation of GLD-360 lightning data and SATCAST convective initiation data into the RAP and RUC models and lightning forecast guidance from nested HRRR runs

Wednesday, 7 November 2012
Symphony III and Foyer (Loews Vanderbilt Hotel)
Tracy Lorraine Smith, NOAA/ESRL/GSD and CIRA, Boulder, CO; and S. S. Weygandt, S. G. Benjamin, C. Alexander, M. Hu, H. Lin, E. W. McCaul Jr., and J. Mecikalski

Two new observation types are being assimilated into real-time parallel versions of the Rapid Refresh (RAP) and Rapid Update Cycle (RUC) mesoscale model systems. The data types are GLD-360 lightning data, which has excellent long range coverage (very helpful for identifying convection over oceanic and other data sparse regions), and SATCAST convective initiation fields, which can help reduce the time-lag for convection initiation in the model forecast.

The RAP is an hourly assimilation system developed at NOAA/ESRL and was implemented at NCEP as a NOAA operational model on May 1, 2012. The RAP replaced the RUC at NCEP, but a real-time version of the RUC continues to run in a research mode at ESRL. The 3-km High Resolution Rapid Refresh (HRRR) runs hourly out to 15 hours as a nest within the ESLR real-time experimental RAP. The RAP and HRRR both use the WRF ARW model core and the Gridpoint Statistical Interpolation (GSI) is used within an hourly cycle to assimilate a wide variety of observations (including radar data) to initialize the RAP.

The GLD360 lightning data are assimilated by converting the flash rate density (per RAP grid box over a 40 min. period) to a proxy reflectivity field, which is then used within the diabatic digital filter initiation (DFI) to specify a latent heating based temperature tendency. This temperature tendency then induces storm-scale convergent and divergent winds. The extended range coverage of the GLD-360 lightning data is also very helpful for initializing oceanic convection in the RAP. The McCaul lightning diagnostic algorithm is applied to HRRR output fields to estimate lightning potential as a function of graupel flux and vertically integrated ice content.

At present, the assimilation of SATCAST (SATellite Convection AnalySis and Tracking) data into the RUC use eight satellite-based predictors to form a convective index (CI), which is then translated into a radar reflectivity proxy value and used in a manner similar to that for the lightning assimilation. Work is ongoing to 1) port the SATCAST assimilation from the RUC to the RAP model and 2) evolve the SATCAST input fields to be more closely related to the magnitude of the cloud-top cooling rates observed by the satellite.

At the conference, we will report on latest results in all of these areas. For the lightning and SATCAST data assimilation portions, we will focus on case study analysis of the forecast impact as well as objective verification. Evaluation of the impact on RAP and HRRR from assimilation at these data at 13-km (within the RAP) will be complemented by discussion of initial results from 3-km assimilation tests. Due to the small-scale nature of these assimilation inputs, we expect potentially greater forecast improvement form use of these data at 3-km. For the lightning diagnostic work, we will focus on case study evaluation of results.

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner