SAFARI: Estimation and assessment of the marine atmospheric boundary layer stratification using synthetic aperture radar imagery
Abstract
Datasets describing the marine atmospheric boundary layer (MABL ~1-5 km) and the exchanges of water vapor and heat between the ocean and atmosphere across large geographic regions are sparse. We propose to use freely available high-resolution, O(10 m), wide-swath, O(300 km), satellite synthetic aperture radar (SAR) images to characterize the state of the MABL and the air-sea interface stratification, which is a key input to all air-sea flux parameterizations. SAR is an effective tool in this endeavor because coherent structure-induced modulations of the surface wind introduce local variations in the surface wind stress. Centimeter-scale capillary and small gravity ocean surface waves respond sensitively and rapidly to the local surface stress and this surface wave field variability is readily detectable by SAR because the backscatter signal has a strong contribution from Bragg scattering. Advanced signal processing of the radar backscatter from polar-orbiting satellites, like Sentinel-1 (S-1), produces almost instantaneous maps of ocean surface roughness that include the surface imprints of the MABL coherent structures. SAR can see through cloud cover and measure the imprint of the atmospheric wind stress and scales larger than 800 m can be described by three dominant categories: 1) unstable stratification (cells or convection), 2) near-neutral stratification (rolls or wind streaks), or 3) stable stratification (lack of rolls or cells). These flow patterns influence the movement of heat, moisture, and winds within the lower atmosphere. The coherent structures inducing the wind stress patterns uniquely map into air-sea stratification classes.We propose to develop an automatic detection system for the S-1 width swath imaging modes (e.g. interferometric wide swath IW). S-1 wide swath imagery typically covers large 250x(>)400 km areas. To effectively utilize the rich information embedded in these images, segmentation detection models need to be trained and verified specifically for the S-1 imagery. This technique can provide high-resolution spatial information about atmosphericstratification. Associated field and modeling experiments in the SAFARI-DRI could likely (co-location dependent) provide key information to extract quantitative MABL data from SAR imagery. Low-level clouds especially across the tropical ocean represent one of thedominant large-scale feedback terms in the Earth s radiation budget. Recent research in the tropical NW Atlantic suggests that one model weakness may be reliance on predicting cloud variability using local thermodynamic state variables without consideration of mesoscale changes (100-1000 km) in the coupled ocean-atmosphere. This project seeks to investigate connections between the SAR-derivedinformation from wide-swath imagery and low-level clouds under the varying mesoscale organization. Consequently, the guiding project questions are: 1) Do time/space shifts between the MABL large-scale eddy field (cell-to-roll) at the mesoscale help explain the varying low-level cloud fraction? 2) How does wind field variability impact both processes and perhaps their coupling? The project objectives are then:Develop machine learning methods to extract atmospheric stratification state information from S-1 wide swath imagery. Validate/verify the model performance with external information from in-situ observations, other remotely sensed datasets, or model output.Address how the SAR surface textures are related to low-level clouds.Analyze how SAR-sensed inhomogeneities like cold pools or rain cells are related to mesoscale cloud organization or wind variability.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Nov 09, 2024
- Source ID
- N000142412760
Entities
People
- Justin Stopa
Organizations
- Office of Naval Research
- United States Navy
- University of Hawaiʻi System