During the IHOP (International H2O Project) the NOAA Forecast Systems Laboratory will be running several configurations of the RUC and MM5 models over various domains in part to provide forecast and nowcast guidance for project operations. IHOP is scheduled to run from 13 May to 25 June over the Southern Great Plains. The main goal of IHOP is improved characterization of the four-dimensional (4-D) distribution of water vapor and its application to improving the understanding and prediction of convection.

The FSL experimental models planned for IHOP include the RUC model at 20 km and 10 km horizontal grid resolution, the WRF model initialized with the RUC analysis at 10 km horizontal grid resolution, the MM5hot model initialized with LAPS at 12 km and 4 km horizontal grid resolution, and possibly the WRF model initialized with LAPS at 12 km and 4 km horizontal grid resolution. This extensive array of real-time models, in addition to a number of other experimental and operational models, will be available for inspection during IHOP, providing an unprecedented real-time opportunity to study the capabilities of models at these resolutions in predicting what is climatologically in springtime a very wide variety of convective activity. In addition, the authors will be participating in IHOP as forecasters or nowcasters for real-time project support during IHOP, and will be working closely with SPC personnel who will be preparing longer range IHOP forecasts. This will allow for a preliminary investigation as to how best to use such models for forecast guidance.

The emphasis of this paper will be on a detailed subjective evaluation of model performance, particularly in ability to predict pre-convective fields such as CAPE, CIN, and vertical wind shear, as well as low-level boundaries, and convective development. Of importance for IHOP will be convective initiation, as well as the type and behavior of convection once initiated. During the last Severe Local Storms Conference one of the authors presented a couple of intriguingly accurate real-time supercell simulations from our MM5 10 km model initialized with LAPS over eastern Colorado, and a comparison to the same model rerun at 2 km horizontal grid resolution. We look forward to seeing whether such forecasts will be as successful in an environment that should present more frequent opportunities to predict such storms, as well as a variety of other convection.