Point Cloud Technologies for Geomorphologists

The first chapter is an introduction to the book. It explains what is meant by pointcloud through the manuscript as well as how the technology has emerged through the field of geomorphology. Within this framework, the laser and photogrammetric technologies (airborne laser scanner, terrestrial laser scanner and structure from motion multiple view stereophototogrammetry) used through the book are briefly introduced, as well as how these methods have come to combine with and complete traditional geodesy. The first chapter will also be the stepping stone to the discussion, which is further detailed in Chap. 7. It explains how these new technologies need to be accompanied with paradigm modifications and how the single-point measurement to the point-cloud has generated a shift in how geomorphologists work in the field and in the virtual space. From this introduction–that includes both the technological aspects and the conceptual aspects further developed in the book–the chapter ends with the structure of the other chapters and how they logically articulate with one another.

The second chapter presents the generation process of pointcloud in geomorphology. The chapter starts with laser technologies that have given rise to aerial laser scanner (ALS) and terrestrial laser scanners (TLS) and how they have emerged in geomorphology, notably through the needs in fluvial geomorphology to detect minute topographic changes, and so over long channel stretches. The characteristics of lasers and the different types of laser solutions are presented with the RIEGL sensors and one of the Leica sensors. Then follows a more practical section explaining how to collect your pointcloud and how to generate it. This first section on laser technologies is then followed by a section on SfM-MVS (structure-from-motion–multiple-view-stereophotogrammetry), with a section on the emergence of SfM-MVS in geosciences and a practical section on how to collect data in the field and the processing steps in Metashape^® to generate the pointcloud data. The chapter ends with “photogrammology”, which is a non-quantitative use of SfM-MVS in order to retrieve shapes or stitch images, when otherwise necessary information is not available. The chapter includes worked examples.

This chapter starts with a presentation of the common pointcloud formats used in geomorphology and civil engineering, as well as a rapid presentation of three freeware and software (with free academic licence) available to read and process pointcloud data. The chapter then explains the registration process of pointclouds and the classification of points including some of the most commonly used algorithms (slope-based and modified slope-based filters, the progressive morphological filter, the low-topographic gradient filter). The chapter then presents some new ideas for processing pointclouds in geomorphology and how data can be locally vectorized to decrease the processing time. After this first section on “one pointcloud”, the chapter then rapidly presents a comparison of two pointclouds running in the CloudCompare software (this is a topic that has been well-developed by other geomorphologists and thus the section is not developed further). The chapter then continues on the conversion of data format. As numerous geomorphologists will work in a GIS environment that deals with shapefiles and gridded data, the chapter then discusses the limitations and advantages of converting data formats (TIN and grids) before presenting the common measurements and derivatives that can be taken from converted pointclouds. This section also has worked examples using the R language and the LidR library to extract gridded data from pointclouds, etc. The chapter is then finished with quality measurement and error analysis of the pointclouds, including RMSE, MBE, MAE, SDE and SE, with once again worked examples in R.

This chapter starts the section of applications of pointcloud technologies in geomorphology, and as this section starts at the shore, going up the mountain. This chapter presents the advantages and difficulties of working and collecting pointclouds in coastal and floodplain areas. The chapter starts with some very basic ideas on river catchments, sediment transport and erosion creating river landscapes, as well as the main processes modifying the coasts. This first section acts as a rapid presentation to non-geomorphologists before presenting the pointcloud technologies and the floodplains, with a space division at different scales showing how different landforms need to be approached with different data acquisition methods. The same division of space by scales and landforms has been then applied to coastal landforms, and advantages and disadvantages of the different methods are also presented. The chapter then ends with two sets of worked examples using a meander of the Alabama River (USA) and a coastal example in Cuckmere Haven in the UK.

This chapter continues the progression upstream, and it then presents the use of pointcloud technologies for hillslopes. The chapter starts with a short overview of some of the slope processes and their numerical expression and how those concepts can be used for the application of geomorphology to hazard management. The chapter then presents a set of applications of pointcloud technologies to hillslope landforms characterized by different scales. As the scales vary, they present different challenges both in term of data acquisition and processing. The slopes are thus divided between interfluves, the drainage network and the mass movements. The chapter ends with a few worked examples using LiDAR and SfM-MVS data, including cleaning the dataset from vegetation, using the vegetation and trees as a proxy of surface processes.

This chapter turns towards a specific type of hillslopes: volcanoes and how pointclouds can be used in volcanic geomorphology. The chapter first presents general features of volcanic geomorphology, before presenting the use of SfM-MVS and LiDAR to deal with the specific difficulties of working in volcanic environment (i.e. recording steep craters’ slopes or staying away from an erupting vent …). This chapter also uses this opportunity to present how pointcloud technologies can be used for micro-geomorphology (in the present case, from the example of a pyroclastic flow deposit surface differential erosion) as well as micro-geomorphology in the laboratory (in the example given, a small sand volcanoes). From the examples, the chapter also proposes a short discussion of the role of grid cell size and sampling strategy from a volcanic geomorphology dataset processed in R.

The last chapter builds on the previous 6 chapters, which are more technical, to propose a reflection of what it means for the geomorphologist to work with high-resolution pointcloud technology. First, the chapter explains that values such as sediment density and compaction, which could be safely ignored in the past are now having an increasing importance in the measured changes, because we can access minute variation, which have more origins than simply erosion and/or deposition. Secondly, the chapter explains that pointcloud technologies have generated a transfer of the geomorphological act, where instead of engaging a reflection leading to a measure in the field, pointcloud technologies are now bringing the geomorphological act in the virtual space. It is now in front of the computer that the geomorphologist will eventually chose a point or extract a transect or generate a 3D volume. The field can be fully automated, robotic almost. Finally, the chapter also presents new fields of opportunities that the geomorphologist can open with pointcloud technologies, like measuring density and density variations, compaction and decompaction occurring with Earth surface processes. These otherwise punctual measures that were limited to civil engineering use can now make their way to geomorphology as well. As a closing chapter, one can see that pointcloud technology is not only a technological change, it has modified the geomorphological act as well as its prospect and assumptions and paradigms.