Application Of Airborne LiDAR(Light Detection And Ranging)for Quantitative Tectonic Geomorphology

Over the past few decades, the Light Detection and Ranging (LiDAR) technique has gained popularity for its high resolution, high precision and efficient data collecting ability in many fields. GPS, IMU, laser scanner and aerial camera are integrated into one unique platform, cooperatively working as one system. As a sensor technology, the LiDAR system includes active scanning pattern (laser scanner) and passive electricity imaging pattern (aerial camera). The laser scanner continuously emits laser pulses, and then receives multiple reflection signals and calculates the distances between the lens and scanned objects. Furthermore, the slope, the roughness or the reflectivity of the object surface can also be generated from the original signals. On the other hand, the camera captures the real time true color images of the objects surface. Finally, point cloud and aerial photos along the scanning pathway are fused to generate a three dimensional digital model of the scanned object. With respect to the traditional remote sensing methods, the LiDAR technique is regarded as a technical revolution in the photogrammetric field. Geoscientists would benefit from the high density point cloud which provides a high～resolution and high～precision terrain model. Additionally, plants or artificial buildings on the surface can be partly or even mostly removed from the original point cloud, so that the true object is imaged. Until now, the LiDAR system has been widely employed in geosciences, such as basic surveying and mapping, cyber city buliding, forestry investigation, and railway and power line prospecting. For different applications, LiDAR device can be installed on the satellite, aircraft, vehicle or tripod on the ground. Among all the platforms, the airborne LiDAR system is known as a convenient and efficient active remote sensing method with short operation time, high data precision and weatherproof availability.It is easy to understand that the airborne LiDAR doesn’t exactly work like the traditional geodetic and photogrammetric ways. To express clearly and conveniently, this dissertation first introduces the airborne LiDAR system components, its working principle, sources of errors and post～processing algorithms. Next, the latest developments and significant projects about the airborne LiDAR applications in geosciences, especially in the seismogeology are reviewed. Taking the Haiyuan airborne LiDAR scanning project as an example, a left～lateral offset channel deeply incised into the alluvial fan of the Shaomayin creek is used to demonstrate how to measure the horizontal coseismic and cumulative offsets from the high resolution digital elevation models(DEMs) derived from point cloud. Following the same method, more than250channel offsets are measured along the fault strike, which exhibit the potential applications of the LiDAR technique in fault activity studies. At last, the LiDAR data collected after6months of the1999Hector Mine earthquake are revisited. The coseismic slip distribution measured from LiDAR data and field work are comprehensively compared, which could provide a distinct chance to evaluate the advantages and disadvantages of the LiDAR technique in research of neotectonics.Chapter One describes the main components of the LiDAR system and its basic working principle. At the beginning, the development of LiDAR technique since1970U.S. Apollo Project using the laser ranging system is reviewed. Then, the basic principle and the technical characteristics of the airborne LiDAR system are briefly illustrated. To introduce the latest development and main application fields, some special airborne LiDAR systems such as Scanning Hydrographic Operational Airborne LiDAR Survey (U.S. Navy), Laser Vegetation Imaging Sensor (NASA) and Ice Cloud and Land Elevation Satellite (NASA) are presented. Next, the working mechanism of each major component and key technique, such as the laser ranging subsystem, dynamic GPS position unit, inertial measurement unit, and multiple antennas measurement array is depicted one by one. Then, the geometric model of LiDAR surveying is derived from the laser ranging principle and the scanning mode. Several essential concepts are also introduced for better understanding of the model derivation process. At last, the differences between the airborne LiDAR and the classical photogrammetric system, between LiDAR and Interferometric Synthetic Aperture Radar (InSAR) are discussed to indicate the strengths and weaknesses of the LiDAR technique. Furthermore, the whole data producing workflow is described and the critical specifications of the most popular equipment in the business market are listed.Chapter Two consists of three parts, which include the types and formats of LiDAR data products, the sources of data error and the post～processing algorithms. Except for the directly collecting data such as massive discrete points, successive full waveform recordings and digital aerial images, some indirectly collected data. i.e. the interpolated regular grid models (DEM or DSM), are also called as LiDAR data products. For the most original and maybe the most important point cloud dataset, the source of error is too complicated to be removed only just depending on the mathematical modeling method. In fact, the system calibration and error models are applied to weaken the influence of system error. As to the final data quality, the traditional ground control point (GCP) method cannot be directly applied because of the discreteness of point cloud. Especially for the horizontal accuracy evaluation, the pre～laying regular artificial objects with high reflectivity will be helpful for the precision evaluation. Over the past dozen years, the filtering and classification algorithms have been the investigation focus for all LiDAR application fields. However, no one method can achieve the goal of fully automatic data processing until now. When the topography in the real world is sampled to hundreds of millions points, semiautomatic software which needs a great deal of manpower and time would limit further LiDAR applications seriously.Chapter Three is divided into two parts. The first section focuses on the representative applications of LiDAR technique in geosciences. For example, the research of the thickness of the polar ice caps and global climate change benefit from the laser altimetry loaded by the ICESat satellite. The similar examples include the migration detection of the shoreline by fusion of the LiDAR data and the tidal recordings, the slip mechanism study of the landslide material based on the repeated high～precision LiDAR scanning, the quantitative tectonic geomorphology research and numerical simulation of topographic dynamic process or the future tendency prediction. Without any questions, all examples indicate the potentials of the LiDAR technique in geosciences. The second section illustrates several kinds of significant LiDAR scanning sample projects at home and abroad, such as the San Andreas fault LiDAR scanning project (B4), the quick response and damage assessment projects after several large earthquakes (i.e. Haiti M7.0earthquake, New Zealand Darfield M7.1earthquake and Wenchuang M8.0earthquake) and the Heihe Watershed Allied Telemetry Experimental Research (HiWATER).Chapter Four takes the Haiyuan fault LiDAR scanning project as an example to demonstrate the whole workflow of the LiDAR data production, including how to design the critical parameters based on the special scientific target and how to evaluate the final data quality by using statistic GPS measurements. With the development of the LiDAR technique, the large～scale high resolution DEMs would play an important role in the identification of active faults and the establishment of the slip distribution along fault strikes. So an active fault can be mapped even at the unprecedented scale of1:1000, which has great significance for active fault studies especially in urban regions. As precious experiences proved, the new phenomena would result from the high resolution topography model derived from the point cloud, and then the new knowledge would be obtained from the observed geologic and geomorphic phenomena. Part of this chapter has been published in the Chinese Science Bulletin volume58issue1(2013).Chapter Five presents a detailed analysis on the offset measurement of one left～lateral offset channel incised into the alluvial fan of the Shaomayin creek by using LaDicaoz, one tool developed in the Matlab environment. LiDAR provides a totally new approach to obtain high quality DEMs effectively. This work takes the Haiyuan fault in Gansu Province as an example to show how LiDAR data can be used to improve the study of active faults and the risk assessment of related hazards. In the eastern segment of the Haiyuan fault, the Shaomayin site has been comprehensively investigated in previous research because of its exemplary tectonic topographic features. Based on unprecedented LiDAR data, the horizontal and vertical coseismic offsets at the Shaomayin site are described. The measured horizontal value is about8.6m, and the vertical value is about0.8m. Using prior dating ages sampled from the same location, this work estimates the horizontal slip rate as4.0±1.0mm/a with high confidence and define that the lower bound of the vertical slip rate is0.4±0.1mm/a since the Holocene. LiDAR data can repeat the measurements of field work on quantifying offsets of tectonic landform features quite well. The offset landforms are visualized on an office computer workstation easily, and specialized software can be used to obtain displacement quantitatively. By combining precious chronological results, the fundamental link between fault activity and large earthquakes is better recognized, and the potential risk of future earthquakes is estimated.Chapter Six puts attention to the analysis of the slip measurements along the Laohushan fault and explores the further application of the LiDAR dataset in active fault studies. The Laohushan fault is the middle～east segment of the Haiyuan fault which accommodates the eastward component of movement between Tibet and the Gobi Ala Shan platform. In the1990s, the researchers mapped the fault at a scale of1:50000. Based on the previous work and the high resolution DEMs,225channel (including ridges and side slopes) offsets are carefully measured at203locations sequentially. The offsets less than20m are picked out and further investigated, which can be used to establish the relatively reliable coseismic slip distribution and rupture range of the1888Jingtai earthquake. According to the transform formula between the moment magnitude and the seismic moment proposed by Kanamonri, the magnitude of the1888Jingtai earthquake is modified to be Mw6.9instead of the original magnitude of Mw6.25. Additionally, the moment magnitude and the surface range of the earthquakes prior to1888are also generally sketched based on the Songshan paleoseismic trench and the cumulative slip distribution derived from LiDAR data. So it is proposed that the earthquake recurrence pattern of the Tianzhu seismic gap should follow the segmented patch rupture mode. Given the hypothesis of millennial recurrence time, the long term slip rate of the Laohushan fault is estimated to be6～7mm/a.Chapter Seven revisits the1999Hector Mine earthquake which is regarded as the first LiDAR scanning project with the clear scientific target for earthquake research. This thesis examines the details of the along～fault slip distribution of this earthquake based on255horizontal and85vertical displacement measurements using a0.5m DEM derived from the LiDAR imagery. The measurements based on LiDAR dataset are highest in the epicentral region, and taper in both directions, consistent with earlier findings by other workers. The maximum dextral displacement measured from LiDAR imagery is6.60±1.10m, located about700m south of the highest field measurement (5.25±0.85m).This work also illustrates the difficulty in acquiring displacements smaller than1m using LiDAR imagery alone. The slip variation is analyzed to see if it is affected by rock type, and whether variations are statistically significant. This study demonstrates that a post～earthquake airborne LiDAR survey can produce an along fault horizontal and vertical offset distribution plot with a quality comparable to a field survey. While LiDAR data can provide higher sampling density and enable rapid data analysis for documenting slip distributions, this work finds that, relative to field methods, it has a limited ability to quantitatively document distributed and diffuse deformation. This work recommends a combined approach that merges field observation with LiDAR analysis, so that the best attributes of both quantitative topographic and geological insight are utilized in concert to make best estimates of offsets and their uncertaintiesIn the dissertation, the main work is based on the new technique and new phenomena from the unprecedented high quality dataset. To demonstrate the airborne LiDAR potential and powerful application in neotectonics, the hardware components of the airborne LiDAR system, the popular algorithm of the software, the kinds of data product type, the mathematics model and the source of the error have been briefly introduced. Furthermore, the Haiyuan LiDAR scanning project is taken as an example to illustrate the whole workflow. One channel incised into the alluvial fan of the Shaomayin creek is picked out to demonstrate how to measure the single channel offset. Then, the coseismic and cumulative offsets along the Laohushan fault are measured following the same method. The offset measurements along the fault strike are analyzed, and then combined with the previous geologic mapping and chronology results to establish the paleoearthquake surface range and estimate the long term slip rate, which could help understand the fault behavior and obtain the new knowledge about the fault activities.The innovation of this paper can be summarized to four points.(1) Exploring the new technique. The airborne LiDAR is an low cost and high efficiency remote sensing technique. Over the past decades, it has been successfully applied in neotectonics since the1999Hector Mine earthquake. However, it is still a long way to the target of its wide application for active fault study in China. The implementation of the Haiyuan fault LiDAR scanning project would be thought as an exploration to demonstrate the whole workflow of the LiDAR data production, which consists of the route planning, point cloud data collection, post～process analysis and final data quality evaluation. The Haiyuan LiDAR scanning project is supposed to provide significant technical reference for the future similar LiDAR scanning projects in neotectonics study.(2) Observing new phenomena. The offset channels along the fault strike are so widespread that could be treated as the external geomorphic signals of the fault activities. At the Shaomayin site, the coseismic offset has never been reported in the previous work. However, with the help of the high resolution topography, the horizontal and vertical coseismic offsets are accurately measured. Combining with the age dating data, the average horizontal and vertical slip rate can be estimated.(3) Obtaining the new knowledge. Limited by short time or budget, the traditional field work usually is confined to a few segments or even individual sites. In contrast, more than200offsets are measured from the high resolution topography model derived from LiDAR point cloud along the Laohushan fault, which would provide dense offset measurements to estimate the surface rupture range and slip distribution as accurate as possible. Above results are analyzed to calibrate the1888Jingtai earthquake moment magnitude from Mw6.25to Mw6.9.(4) Expanding the new application. The airborne LiDAR technique has demonstrated its powerful potentials in rapid acquisition of high～resolution tripped topography. However, the air traffic control, expensive device and low degree of automation in data post～processing, limit the further application of the LiDAR technique. The experiences gained from the Haiyuan fault LiDAR scanning project would contribute so much in theories and methods on how to apply the LiDAR technique in active fault studies. The subsequent researches, such as active fault mapping, geological mapping and fault activity study should be carried out as quickly as possible.