Evaluating A Blending Algorithm for Atmospheric Refractivity Using CASPER Measurements

Evaporation ducts are one of the most frequently encountered radar ducts in the marine atmosphere. They occur due to the close interaction between the atmosphere and the underlying ocean as a result of heat and water vapor exchange. The vertical profiles of refractivity are generally obtained using a Monin-Obukhov Similarity Theory (MOST) based 1-D model such as the Navy’s Atmospheric Vertical Surface Layer Model (NAVSLaM), when only bulk parameters at a single level are observed or modeled in the surface layer.

When simulating the effects of the atmosphere on radio frequency wave propagations, the refractivity profiles from the entire air column is needed, including that from the atmospheric surface layer, or the evaporation ducts. Yet, the refractivity profiles from forecast models or from rawinsonde measurements normally are not capable of representing accurate evaporation duct properties. One general practice in modeling radio frequency (RF) propagation in the atmosphere, using weather forecast model results, is to pad a surface layer refractivity profile to the model-derived refractivity profile for obtaining a complete refractivity profile from the surface to the top of the atmosphere.

A similar requirement exists when using refractivity profiles from rawinsonde observations for simulating propagation through the measured atmosphere. For sounding measurements obtained from rawinsonde launches at sea, the lowest levels are also problematic. These lowest levels, up to about 50 m above the surface, is normally affected by flow distortion and thermal effect induced by the ship. So the lowest levels from the sounding are not ideal for quantifying evaporation ducts. The lowest 10s of meters in the measured refractivity profiles need to come from evaporation duct models and to be blended to the observed sounding profile.

We developed a ‘blending’ technique that can be used to extend the measured or model refractivity profiles down to the surface layer to include the evaporation duct. The algorithm was COAMPS running in the single-column model, with high vertical resolution and thus capable of resolving the evaporation duct. We also modified the boundary layer mixing portion of COAMPS boundary layer parameterization so that the model can represent the surface layer mixing following MOST. This method is referred to as the Single Column Model (SCM) Blending Algorithm (SCMBA). As with any boundary layer model, large-scale forcing needs to be specified, which is not always available. Here we ran SCMBA, using a nudging technique, that nudged the SCM to follow the modeled or measured profiles above a designated level above the MASL. In this manner, the boundary layer and surface layer evolved forward in time, within the model, to which the turbulent mixing shaped the surface layer and resulting upper boundary layer gradually moved towards the modeled or measured profiles.

In this presentation, results from the new blending scheme will be compared to observed refractivity profiles from tethered balloon-based measurements on board small boats, and from measurements on the FLIP boom with a sensor system reaching to the surface during the Coupled Air-Sea Processes and Electromagnetic ducting Research (CASPER) project. The small boat measurements provide the first opportunities to evaluate any evaporation model using in-situ measurements. The FLIP based measurements also extend the measurements consistently to the immediate surface adding more data as evaluation dataset. We will also apply SCMBA to COAMPS profiles during the CASPER field measurements and compare the results with the currently operational approach.