The Assessment and Validation of Atmospheric Water Vapor in Global Analysis and Forecasts Accross the Broader Caribbean

Water vapor has a fundamental role in numerical weather analyses and forecasts. The transport of water vapor within a model serves as a primary method in which energy in redistributed within the model domain. The Caribbean is a region where model analysis are thought to have anomalously large errors in water vapor. The Continuously Operating Caribbean Observational Network (COCONet) is a collaborative project to create an international network of Global Navigation Satellite System (GNSS) stations in the Caribbean for natural hazards research. Atmospheric data products generated from COCONet include estimates of column integrated tropospheric water vapor, precipitation, as well as measurements of surface temperature, relative humidity, pressure, and horizontal winds. The distribution of COCONet stations across the circum Caribbean allows for a robust evaluation of existing atmospheric analysis and forecast products. Previous research indicates that current atmospheric analysis models have biases in precipitable water vapor (PWV) that are dependent on the magnitude of PWV. Analysis conditions appear too moist in relatively low PWV (<40 mm) conditions, and too dry when there is relatively large PWV (>40 mm). This presentation will focus on the comparison of COCONet derived atmospheric observations with GFS forecasts fields extending out to 168 hours. Specific emphasis will be given to the hurricane season, where water vapor products are especially valuable.