G13B-0804 1340h
Mapping Landforms in Geomorphology Using SRTM Data
Identification and mapping of landforms in geomorphology is based on geological and geomorphological survey and interpretation of topographic maps and aerial photos. The traditionally techniques are affected by a high degree of uncertainty. In order to reduce this type of errors new analysis processes calibrated elaborations on Digital Terrain Models (DTM). To this end, we have used SRTM (Shuttle Radar Topography Mission) digital elevation data for Italy. To investigate the question of model sensitivity to various remote-sensing techniques for scaling or aggregation of landscape attributes, it is necessary to work within the context of a given model's data requirements and sensitivity. Model sensitivity to input data error propagation can be evaluated to specify the form and acceptable limits of accuracy of input data sets describing land surface attributes. For certain geomorphological processes that are strongly dependent on tectonic evolution in a certain areas, much of the process variations at sufficiently large spatial and temporal scales, can be explained with direct observations of geometric system using landscape satellite. Our method of analysis is based on a delineation of surface landscape attribute using a multipixel delineation algorithm. The algorithm is designed to map, on a regional scale, surface drainage patterns that may include diffuse flow, curvature and relief analysis. Some landforms (e.g. alluvial fans) show a close correspondence between their limits (curvature and relief), and the direction and diffusion of hydrological drainage pattern, so this type of analysis can represent an objective method to recognize these landforms and to understand their evolution (e.g. telescopic alluvial fan depending on normal faulting). We present a preliminary analysis of SRTM data from Umbria Region. The algorithm is a significant improvement and images of model surface flow provide an excellent means of using SRTM data to study interactions between faulting and drainage patterns. As an example, we discuss some observations of alluvial fans formations induced by the propagation of the trust fault system.
G13B-0805 1340h
An Evaluation of a Spline based Vegetation Removal Algorithm for Airborne Lidar Measurements
Utilizing fast pulse rate laser measuring devices, airborne lidar technique provides an efficient method for the collection of three dimensional point coordinates. Because the lidar measured points include both the surface of vegetation and ground measurements, the generation of digital elevation model for the ground requires further processing. This study utilizes a spline based filtering algorithm. The ISPRS test data set, which includes fifteen datasets collected from eight different areas, is utilized for the evaluation. The result is compared with the reference data. The characteristics of those mistakenly classified points are investigated, both of the commission and omission errors. It is found that the success rate ranges from 83% to 99%, depending on the terrain types.
G13B-0806 1340h
Distinguishing process through form in alluvial fans, eastern Death Valley, California
We use high resolution (1m) ALSM data to assess the morphology as a function of length-scale of alluvial fan form in Eastern Death Valley, California. The goal of the study was to develop topographic fingerprints that would enable formative processes to be distinguished from form alone. Thus, alluvial fans were divided into two categories based on field observations of the sedimentological features and apparent formative processes: Clast Rich, predominately debris flow (CR), and Mixed Flow, largely a mixture of debris flow and fluvial processes (MF). The data used in the study were collected and post-processed into a bare-earth model by personnel from the National Center for Airborne Laser Mapping (NCALM). Gridded DEMs of the bare-earth data were generated in ESRI's ArcInfo, from which individual alluvial fans were delineated via 1 m hillshade and curvature grids. DEMs of individual fans were detrended by removing a second-order polynomial surface from the original fan surface. The detrended fan surface revealed topography at relatively short length-scales and removed the dominant longitudinal slope trend that would wash-out the local-relief signature. We examined local-relief as a function of length-scale and used this parameter to determine if statistical differences in form occurred between the CR and MF fans. We find three distinct topographic fingerprints that enable CR fans to be distinguished from MF fans. 1. For any length-scale, local-relief is higher in CR fans than in MF fans; 2. MF fans show a more pronounced positive skewness of local-relief than do CR fans; and 3.The average slope, measured at a length-scale of 1 m, separates CR from MF fans . These fingerprints would be unavailable from a standard 10 m or 30 m DEM. The values reflect differences in surface variability due to varying structures (debris flow lobes, channel, interfluves, etc.) on the fans. Very high-resolution DEM data has the potential to fulfill the long-standing goal that process may be tractable from form. These fingerprints should be useful to modelers because they provide a benchmark to evaluate the success of physical and numerical models.
G13B-0807 1340h
Surface Fractal Characterization of Topography and Landcover Using Airborne Laser Swath Mapping
High density, small footprint Airborne Laser Swath Mapping (ALSM, also referred to as LiDAR) is used to determine surface fractal dimensionality of various types of topography and landcover throughout southeastern Idaho. Surface fractal mapping and classification are used to determine principal length scales found in different types of topography and vegetation. Knowledge of fractal dimensions can assist in determining sampling intervals, error statistics, and vegetation removal when creating digital terrain models (DTMs) and vegetation height maps. Terrain types investigated in this study include semi-arid rangelands, forests, riparian areas, canyons, landslides, and agricultural regions.
G13B-0808 1340h
An Alluvial Fan Spectral Surface Analysis Comparison of Airborne LIDAR and GPS Data for the Calico Fault, Mojave Desert, California
Airborne LIDAR imaging is an effective tool to acquire large areas of high-resolution topography for surface analysis. Surface roughness characteristics can be matched throughout the LIDAR coverage area in order to determine relative surface chronologies, which will aid in the estimation of fault slip rates. It is then possible to "spot check" surfaces with absolute dating methods. Un-vegetated desert alluvial fan surfaces of differing ages and roughness in the Mojave Desert, California, provide an attractive natural setting to test how well surface characteristics are represented by LIDAR data in comparison to ground survey points. LIDAR data were acquired over a chronosequence of geomorphic surfaces at the foot of the northern Rodman Mountains as part of a companion study of the slip rate of the Calico fault. On the ground, high-resolution differential GPS data were measured from four separate fan surfaces of different ages and smoothness characteristics. GPS survey points were acquired at 25 cm spacing on orthogonal transects spaced 50 m apart. Spectral analysis of surface elevations provides a tool for comparing the topography signal between LIDAR and survey data. The LIDAR data has an approximate resolution of 1 meter, which is good for detecting landforms, such as bar and swale, on a spectral order of tens of meters. However, the raw LIDAR data points are not the same distance apart in geographic space, so when spectral analysis is applied to the LIDAR data it must be resampled. Alternatively, raw LIDAR data series may be treated as a time series, however this restricts analysis to the swath orientation.
G13B-0809 1340h
Assessing the Ability of Laser Altimeter Return Waveforms To Detect Surface Topographic Change
The capability of medium/large footprint (10 m or greater) full-waveform laser altimeters to penetrate beneath dense vegetation to directly measure sub-canopy topography provides a unique capability for sensing topographic change in the presence of vegetation. In this study, we assess the ability of geolocated laser altimeter return waveforms instead of individual elevation measurements to measure vertical elevation change. The method we use (dubbed the return pulse correlation method) maximizes the shape similarity of near-coincident, vertically-geolocated laser return waveforms from two observation epochs as they are vertically-shifted relative to each other. Using waveform data collected by NASA's Laser Vegetation Imaging Sensor (LVIS) we assess the inherent precision of elevation change measurements derived using the return pulse correlation method. Data collected in 1999, 2003 and 2004 are compared using the pulse correlation method and by simple differencing of coincident elevation measurements. Data collected under "bare earth" conditions as well as in the presence of dense, growing vegetation are utilized, demonstrating the usefulness of this technique for complementing and extending earth surface dynamics measurements made using other remote sensing techniques.
G13B-0810 1340h
Exploiting LiDAR for Regional Morphologic Correlation and Dating of Wave-cut and Fault-Controlled Landforms
The capability to generate high-resolution Digital Elevation Models (DEMs) from LiDAR data (Light Distance and Ranging, also known as Airborne Laser Swath Mapping, or ALSM) across broad geographic regions provides a new tool for studying landscape response to tectonic deformation. Expanded LiDAR coverage from the Plate Boundary Observatory (PBO) and the National Center for Airborne Laser Mapping (NCALM) offers the prospect of applying these data to a variety of tectonic geomorphic studies. The data volume and point-density of LiDAR allows extensive repetition of profile-based landscape analyses without the need for laborious total station transects. Traditional DEMs, such as the USGS National Elevation Dataset, lack the resolution necessary for these types of analyses. We propose to exploit LiDAR data for landform correlation by conducting profile-based morphologic dating (linear and non-linear diffusion) of fault scarps and marine, lake and fluvial shorelines. The resolution and geographic extent of LiDAR coverage makes broad spatial correlations possible, assuming that controls on the hillslope processes are relatively constant across the region. Due to the high data density, numerous topographic profiles can be extracted from a DEM and analyzed for morphologic age. Correlation from profile to profile can then be established by comparing morphologic age for various landforms in a research area. With calibration, morphologic dating also offers the opportunity to constrain absolute ages of landforms. Once calibrated, landforms across the region can be quickly dated via profile-based analysis of the LiDAR-derived DEM. In addition, morphologic comparison of landforms of known age offers the opportunity to test the role of other constraints, such as aspect, microclimate, and substrate type on landform development by diffusive processes. Tests on synthetic profiles demonstrate the ability of morphologic dating to differentiate landforms of morphologic age 50 m$^{2} from ones of 100 m$^{2}. Application of this technique to recently acquired LiDAR datasets in northern California and in the Basin and Range province illustrates the power of morphologic dating for establishing correlations among regional landforms.
G13B-0811 1340h
LiDAR Imagery of the San Andreas Fault Zone at the Vedanta and Olema Ridge Paleoseismic Trench Sites, Pt. Reyes, CA
At the Vedanta and Olema Ridge paleoseismic trench sites along the San Andreas fault (SAF) in Marin County, we experimented with collecting tripod LiDAR (Light Detection And Ranging) data in order to test its utility in stratigraphic and tectonic geomorphic mapping. To characterized the terrain surface surroundings and within the exposed trench walls, we performed ground-based LiDAR surveys using a portable color sensitive tripod-mounted system. To produce a digital terrain model (DTM) for each site, we used a Riegl Z210i laser-scanner to target the ground and saturate it with point targets at three or more locations around the exposed trench. Local geo-referencing and control points were established using temporary auto reflectors. Using the LiDAR-based terrain model software package, ISite3D, we then merged these scans into a single surface model for each site. The same technique was used to image and process the exposed walls of the trench. We found that using a rotating scanning-laser allows us to very rapidly produce ultra-high resolution and quantitative DTMs for geomorphic analysis of a large (>0.1 km2) area surrounding the trench and that that the DTM can be used to resolve fine scale (<2.5 cm) morphologic features associated with the fault. The ability of the LiDAR to resolve color allows us another tool to investigate subtle variations in the soil structure exposed in the trench wall. By artificially modifying the color with false and enhanced colors, we can visually extract information not readily visible to the eye. At the Olema Ridge trench site, the 1906 trace of the SAF lies at the base of an east-facing scarp that formed as a slice of the ridge has been translated northwestward along the fault. As a means to compare technologies, we collected detailed geomorphic data from the site using both Total Station and LiDAR surveys. The superior coverage of the geospatial data recovered from the LiDAR allows for a more accurate rendering of the microtopography. Using the DTM for the Olema Ridge site, we can determine the probable age of the SAF at this location by palinspastic reconstruction of the surface topography and restoration of the ridge sliver back to its slope position. In addition, through comparison of archival photographs taken after the 1906 earthquake, our detailed tectonic morphologic mapping shows how the 1906 rupture has changed geologically over the past 100 years. We also compare the historical photographs to the subsurface stratigraphic section exposed in a trench. A low sedimentation rate and high rates of bioturbation have largely homogenized the stratigraphic data. We see clear evidence for the 1906 rupture and at least two prior earthquakes are also preserved. At the Vedanta marsh site, a 4-m-deep excavation exposes a section of nearly continuous deposition of organic-rich peat layers interbedded with marsh clay and silt, and fluvial to colluvial gravel deposits. This stratigraphic section contains evidence for approximately ten ground-rupturing earthquakes over the past ~3000 years. We tested the LiDAR system on the vertical exposure at the Vedanta site to see if high resolution geospatial data can capture individual stratigraphic layers, faults, and event horizons that can be interpreted directly from the remotely sensed data. In this presentation, we compare the LiDAR to trench logs mapped by photomosaic and graphical techniques.
G13B-0812 1340h
Spatially-aware Processing of Large Raw LiDAR Data Sets
An ultimate goal of LiDAR (LIght Detection And Ranging) data acquisition is to produce a regularly sampled accurate topographic view of the surface of the Earth. Last-return and inverse-distance weighted sampling of raw LiDAR data do not take into account the non-random distribution of raw data points. While elevation data produced by these methods is of high accuracy, gradients are not well-resolved and aliasing artifacts are produced, especially on low gradient surfaces. Because of the volume of data involved, resampling schemes that take into account the spatial distribution of raw data have been cumbersome to implement. We have developed a resampling method that uses the free open-source PostgresSQL database to store the raw LiDAR data indexed spatially and as its original time series. This database permits rapid access to raw data points via spatial queries. A robust and expedient algorithm has been implemented to produce regularly gridded resampled data with a least squares plane fit regression. This algorithm reduces aliasing artifacts on low gradient surfaces. The algorithm is also a proof-of-concept to show that complex spatially-aware processing of large LiDAR data sets is feasible on a reasonable time scale, and will be the basis for further improvements such as vegetation removal.