G43D-01 INVITED
Determination of Event-Dependent Depth of Closure for the South Florida Atlantic Coast Using Airborne Laser Bathymetry
Depth of closure (DOC) is an important concept in coastal engineering that defines the seaward limit of significant net sediment transport along a wave-dominated sandy beach profile. Few surveys measured the DOC over large areas because traditional methods for measuring DOC are time consuming and cost prohibitive. With a dramatic increase in airborne laser bathymetric data in recent years, it has become possible to measure DOC over many kilometers. Reported here is a new method that identifies the DOC using airborne laser bathymetric (ALB) data. The horizontal location of the DOC was determined by differencing 2004 pre- and post-hurricane airborne laser data sets along Palm Beach, Broward, and Miami-Dade counties. Noise in the ALB data were approximately +/- 0.3 m, thus bathymetric variations greater than +/- 0.3 m were considered significant change. The seaward depth where change was less than +/- 0.3 m was interpreted as the DOC. The measured DOC was compared horizontally and vertically to DOC positions that were calculated based on wave data and to geomorphic units at 1046 locations spaced 100 m along the coastline. Calculated DOC values were on average within 2.8 m vertically to the measured DOC in the northern end of the study area. In the southern segment of the study area, however, the calculated DOC was on average deeper than the measured DOC. Small horizontal differences (90 m, on average) between geomorphic boundaries (rock outcrop, hardgrounds) and measured DOC suggest geologic control south of Hillsboro Inlet. Diabathic channel fields match the measured DOC to the north, with a vertical difference of 0.3 m and a horizontal difference of 161 m, on average. Because diabathic channels are hydrodynamically formed (hydromorphodynamic forms), the northern study area appears to be hydrodynamically controlled. Given the ALB data represent a before and after surface for the 2004 hurricane season, the DOC extracted from ALB data in this study is event-dependent, which is different from a normal DOC based on a typical annual wave condition. Because this study measured an event-dependent DOC where waves generated by two hurricanes were larger than large waves in a typical year, the measured DOC represents a deeper DOC than the normal DOC. It is essential to understand temporal changes of the DOC in order to effectively apply the DOC to various applications. Determination of the DOC from ALB data is an accurate, cost-effective and technologically advanced means of determining DOC over large areas that can range from quiescent phases to large storm events.
G43D-02
Automatic Feature Extraction from Airborne Lidar Measurements to Identify Cross-Shore Morphologies Indicative of Beach Erosion
Airborne lidar data were acquired along St. Augustine Beach, Florida seven times between August 2003 and February 2007. To identify sub-aerial morphologies indicative to beach erosion, the data sets were mined extensively by extracting several morphological features using cross-shore profile sampling. For each profile, the features were grouped into erosion or accretion classes dependent on shoreline change measured and their class-conditional probability density functions (PDFs) estimated via Parzen windowing. PDF separability was ranked using symmetric and normalized measures of probability divergence. The more interclass separation provided by a feature, the stronger the relationship with shoreline change variation and greater its potential as an indicator for erosion or accretion. Over short time periods (>1 month), beach slope and beach width ranked highest by providing the most separation and therefore high potential as indicators for erosion. Over longer time periods (>1 year), deviation-from-trend, which is the shoreline’s deviation from the natural strike of the beach, ranked highest. This is significant in that the pier region’s deviation from the natural trend is believed by coastal researchers to be a strong contributing factor to it being an erosion “hot spotâ€. Furthermore, shoreline deviation appears implicitly within the widely used CERC equation for longshore transport. To test the potential of certain morphologies for predicting where a segment of beach might be more prone to erosion or accretion, a Bayesian classifier was implemented and tested on the data set. The highest ranking features selected by the divergence method outperformed those selected by a simple median metric and the correlation coefficient. Overall, high classification rates were achieved supporting the utility of certain features for erosion monitoring. In addition, an analytical diffusion model fit to the ALSM data was used to simulate spreading rate of a beach nourishment, and results were compared to measured change. The method developed provides a framework to mine high-resolution airborne lidar data over beaches and quantify relationships between alongshore variation in morphology and patterns in erosion or accretion.
G43D-03
Repeat Observations of Sea Surface Topography and Ocean Waves in the Southern California Bight From Scanning Airborne Laser Altimetry
Since 2002, the University of Texas at Austin (UT) and the Scripps Institution of Oceanography have been surveying the southern California shoreline from San Diego to Long Beach using a small-footprint, airborne lidar mapping system. In all, twelve lidar shoreline surveys have been flown to measure the topography of the nearshore waves, beaches and adjoining cliffs. In addition, we have conducted four offshore surveys: flights along the 20m isobath to map the ocean surface from Point La Jolla to Dana Point, a distance of over 60 km. These offshore surveys occurred in May 2002, September 2002, May 2003, and December 2003 and thus represent different tide and sea states. Repeat lidar offshore surveys provide high resolution ocean surface transects that can be compared with gravimetric models of the geoid and satellite altimetry: radar and laser. UT is a participant in the Ice, Cloud and land Elevation Satellite (ICESat) mission. The UT Center for Space Research conducts experiments designed to validate the timing, geolocation, and geometric characteristics of the lidar footprints generated by the Geoscience Laser Altimeter System (GLAS), the principal instrument on ICESat. A comparison of ICESat and TOPEX derived sea surface elevations indicates that ICESat observations of the oceans are influenced by a sea state bias. We discuss the characteristics of airborne and space laser measurements of the ocean surface and the use of small-footprint lidar to estimate an ICESat bias related to wave height.
G43D-04 INVITED
Kinematic GPS positioning of an aircraft and the vertical structure of atmospheric refractivity
During the B4 LIDAR survey each flight of the NCALM aircraft was positioned using many different base stations: instantaneous baseline lengths ranged from less than 1 km to more than 200 km. When we used our kinematic GPS processing software (KARS) in its default mode, the individual trajectory solutions were usually mutually biased in the vertical, particularly when the base stations were located in areas with significant topography. In these cases, relative height biases as large as 20 cm were observed. We found that these biases were caused by errors in atmospheric delay modeling which were systematically organized according to the elevation of the base station (or aircraft). We previously demonstrated how static geodetic analysis of the network of base stations could be used to estimate how zenith delay [ZD] varied with station elevation, and that when the KGPS software was forced to use a ZD(h) model that was consistent with these external measurements, the height biases in the various trajectory solutions almost completely disappeared. This 'external calibration' technique was extremely useful in diagnosing the problem we had encountered, but it is not a good general purpose solution for the problem because (i) it works only in areas with significant topography, and (ii) it involves a rather cumbersome interaction between the static and kinematic GPS analyses. In this talk we present an alternative approach based on direct estimation of zenith delay parameters in the context of the kinematic analysis. We are using this approach to reprocess the entire set of B4 flight trajectories.
G43D-05
Airborne Laser Swath Mapping Imagery for GeoEarthScope
UNAVCO is acquiring Airborne Laser Swath Mapping (a.k.a. airborne LiDAR) imagery for GeoEarthScope, a component of the EarthScope Facility project funded by the National Science Foundation. Guided by the UNAVCO GeoEarthScope LiDAR Working Group, these projects are designed and conducted based on community recommendations with respect to target identification and data collection practices so as to provide the EarthScope community with a rich, high quality data set capable of supporting a wide range of interests and applications. Anticipated applications range from operational, such as assisting with EarthScope instrument siting, to pioneering research in many fields of study including tectonophysics, geomorphology and paleoseismology to name a few. As of September 2007, two GeoEarthScope ALSM projects have been completed: 1) a 1500+ sq km project in northern California that focused on the San Andreas fault and other active structures, and 2) a ~450 sq km project in Death Valley that focused on the Death Valley-Fish Lake Valley fault system. The northern California data were collected during an extensive, highly collaborative field campaign in Spring 2007. ALSM data were collected by the National Center for Airborne Laser Mapping (NCALM) with a new generation Optech Gemini scanner at high pulse rate frequencies, and high rate GPS data were collected by regional networks such as PBO and by campaign systems deployed by a team of personnel from Ohio State University, UNAVCO, the U.S. Geological Survey and student volunteers from local universities. Also for this project, primary GeoEarthScope targets were expanded to include additional targets such as the Hayward fault through supplemental funding contributions from the USGS, the City of Berkeley, and the San Francisco Public Utilities Commission. The northern California dataset complements the previously acquired "B4" ALSM dataset in southern California by overlapping B4 coverage along the creeping section of the San Andreas fault and by using B4 style data collection methods. The Death Valley ALSM data were collected in Fall 2006, also by NCALM, as part of an NSF funded research project led by the University of Southern California (USC). Several additional GeoEarthScope ALSM projects are planned for completion prior to conclusion of the EarthScope MREFC in Fall 2008. Proposed future projects include targets in the Intermountain Seismic Belt (e.g. Wasatch fault), southern and eastern California, the Pacific northwest, and Alaska. Like all EarthScope data, GeoEarthScope ALSM imagery will be freely available to the community. This presentation addresses GeoEarthScope ALSM imagery activities from an overall programmatic and logistical perspective. Projects to date are highlighted and future plans are presented. http://www.unavco.org/geoearthscope
G43D-06
High-Resolution LiDAR Topography of the Plate-Boundary Faults in Northern California
GeoEarthScope acquired more than 1500 square km of airborne LiDAR data in northern California, providing high-resolution topographic data of most of the major strike-slip faults in the region. The coverage includes the San Andreas Fault from its northern end near Shelter Cove to near Parkfield, as well as the Rodgers Creek, Maacama, Calaveras, Green Valley, Paicines, and San Gregorio Faults. The Hayward fault was added with funding provided by the US Geological Survey, the City of Berkeley, and the San Francisco Public Utilities Commission. Data coverage is typically one kilometer in width, centered on the fault. In areas of particular fault complexity the swath width was increased to two kilometers, and in selected areas swath width is as wide as five kilometers. A five-km-wide swath was flown perpendicular to the plate boundary immediately south of Cape Mendocino to capture previously unidentified faults and to understand off-fault deformation associated with the transition zone between the transform margin and the Cascadia subduction zone. The data were collected in conjunction with an intensive GPS campaign designed to improve absolute data accuracy and provide quality control. Data processing to classify the LiDAR point data by return type allows users to filter out vegetation and produce high-resolution DEMs of the ground surface beneath forested regions, revealing geomorphic features along and adjacent to the faults. These data will allow more accurate mapping of fault traces in regions where the vegetation canopy has hampered this effort in the past. In addition, the data provide the opportunity to locate potential sites for detailed paleoseismic studies aimed at providing slip rates and event chronologies. The GeoEarthScope LiDAR data will be made available via an interactive data distribution and processing workflow currently under development.
G43D-07
Use of Geodetic Laser Scanning to Evaluate the Curvature of Bedrock Surfaces in an Investigation of Sheeting Joint Formation
We are using aerial and tripod-mounted geodetic laser scanning (GLS) data, together with photography and large-scale geologic mapping, to investigate the formation of sheeting joints in Yosemite National Park. Sheeting joints are opening-mode fractures that form subparallel to the topography, and over broad areas in Yosemite they define the bedrock surface. Rock slabs bounded by sheeting joints superficially resemble the layers of an onion. Our hypothesis is that sheeting joints form where a tensile stress normal to the topographic surface exists in the shallow subsurface. This condition is met where k2 P22 + k3 P33 > γ cosβ, where k2 and k3 are the principal curvatures of the bedrock surface, P22 and P33 are the corresponding normal stresses parallel to the principal stresses, γ is the unit weight of the rock, and β is the slope angle. Sheeting joints are predicted where at least one of the principal curvatures is sufficiently convex (negative) and the corresponding normal stress is sufficiently compressive (negative). We use aerial GLS data with a vertical resolution of ~10 cm and a point spacing of ~1 m to measure the slope and curvature of the bedrock surface at the scale of a ridge or valley. We use tripod-mounted GLS data with a point spacing of ~5 cm, large-scale geologic mapping, and photographs to detect steps between consecutive sheeting joints, with the step height giving the sheet joint spacing. Outcrops hosting sheeting joints have a stair-step appearance with a distinctive curvature signature: high convex curvature at the top of a step, and high concave curvature at the step bottom. Steps between sheeting joints with a spacing of less than a meter or so are difficult to detect using the aerial GLS data. Apparently the interpolation of aerial data onto a grid, necessary for our curvature codes, and the smoothing of gridded data to filter out trees compromises the value of the aerial GLS data in detecting the step edges, even though the vertical resolution of the GLS data should be quite adequate for measuring the step height. A curvature code that does not require gridded data could help detect step edges. The steps can be detected readily with the tripod-mounted GLS data, large-scale geologic mapping, and photographs, which show shadows cast by the steps. Our analyses to date with all the data sets supports our hypothesis for sheet joint formation.
G43D-08
Feature Classification Benefits of Full-Waveform Airborne Laser Terrain Mapping
We present the feature classification benefits of full-waveform return analysis from airborne laser terrain mapping (ALTM). Through the examination of the characteristics of clusters of return pulses in terms of range, number of returns, scattered amplitude, and scattered width, we are able to discriminate among a number of classes of ground features. The University of Texas Center for Space Research has obtained several distinct full-waveform datasets over several ecosystem types including urban and low-height vegetation environments. The UT ALTM system is equipped with a waveform digitizer that allows the simultaneous recording of conventional first and last return lidar data and the corresponding reflection waveforms at the 25 kHz laser pulse repetition rate. Reflecting targets or surfaces within the return laser path are identified using a Gaussian decomposition of the return waveform. We present the waveform properties from different classes of scatters including man-made and natural targets and draw some conclusions about the potential for unsupervised, untrained classification of objects. Finally, we discuss the applications where full-waveform is likely to most benefit ALTM including the ability to identify ground returns in the presence of understory.