Marine boundary layer characteristics, clouds and air-sea interactions Matching observations and model forecasts to diagnose COAMPS and NEPTUNE weaknesses
Abstract
The crux of our work will be model evaluation based on weather conditions using conditional queries to a relational database that contains matched observation and model data for over 300 locations in the United States, Canada, and near shore. The relational database will build over the course of the project as we ingest observations matched with NEPTUNE and COAMPS model output. We use observations, which have consistent accuracy among locations, to avoid the conundrum of using high uncertainty reanalysis grid values to evaluate forecast models in locations where forecast models perform poorly. Over land and coastal regions, we will use data from operational weather stations, and the lowest 2 km of profiles from National Weather Service upper-air soundings. Over ocean, we will use buoy observations and standard and derived satellite products of cloud characteristics from NASA and NOAA polar and geosynchronous satellites as well as model-derived cloud properties. For model grids with < 5 km spacing we will use the closest grid box and for coarser grids we will linearly interpolate to the observation location. For weather conditions that are prone to recurrent errors, we will use a single column model within NEPTUNE to help identify the sensitivity of those errors to alternate parameterizations and different settings. We will also examine NEPTUNE and COAMPS performance in the context of NOAA GFS and HRRR model errors using the same methods to help distinguish errors in specific parameterizations from inadequately represented processes where all models do poorly.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jan 06, 2021
- Source ID
- N000142112116
Entities
People
- Sandra E. Yuter
Organizations
- North Carolina State University
- Office of Naval Research
- United States Navy