AbstractsEngineering

Depth estimation and frontal imaging via X-band marine radar

by David Andrew Honegger




Institution: Oregon State University
Department: Civil Engineering
Degree: PhD
Year: 2015
Keywords: Marine radar; Radar
Record ID: 2061350
Full text PDF: http://hdl.handle.net/1957/55425


Abstract

This dissertation focuses on two core aspects of remote sensing: (a) interpretation of the remotely sensed data to identify and characterize sea surface features of interest, and (b) the quantitative analysis of previously characterized features to produce robust estimates of geophysical variables. Specifically, these aspects are addressed in the context of the nearshore coastal ocean, utilizing a land-based, imaging X-band marine radar. With regard to the latter aspect (b), a robust, phase based depth estimate algorithm is extended to marine radar imagery with the purpose of assessing both absolute and relative performance against optical video derived depths. Depth estimates at both a nearby (< 1 km) and a moderately distant (2-3 km) beach are accurate to ∼ 0.5 m, with global bias of ∼±0.3 m; these results are comparable to those acquired via optical video, and performance trends with depth for the two sensors appear complementary. Extended to larger (inlet) domains that require the spatial coverage of marine radar imagery, algorithm performance remains at roughly the same level during slack; currents induce an unmodeled Doppler shift that produces wave frequency-and depth-dependent error that should not be neglected. Mitigation and solution strategies are discussed. With regard to the former aspect (a), sharp internal and interfacial gradients in the water column at a highly stratified estuary mouth are identified in marine radar imagery. The identified features include an internal, oblique hydraulic jump, a bottom salinity front, and a surface intrusion front. Confirmation of co-located, sharp current gradients that induce surface roughness modulations similar to those observed were carried out using remotely sensed surface current estimates (via airborne ATI-SAR) for the jump, and using vertical current profiles collected along cross-front transects via research vessel for the bottom and surface fronts. Presence of the frontal structures below the identified backscatter intensity contrast patterns is directly corroborated using salinity, temperature, and current data from bottom-mounted, ship-mounted, and ship-cast sensors, and presence of the internal jump is indirectly inferred using an inviscid two-layer model. Horizontal evolution of the jump and fronts in the estuary mouth is tracked in quasi-continuous marine radar image time series.