Spatial Accuracy and Validation of Topographic and Bathymetric Lidar Data

This session features presentations from experts on determining not just the quantitative, but also the qualitative, accuracy of elevation data sets in both the topographic and bathymetric realms.

10:30 AM – 10:45 AM – Airborne lidar accuracy assessment through rigorous sensor error models

Rigorous physics-based lidar sensor models enable detailed error propogation and the assignment of 3D geolocation accuracy predictions to each point in a data product point cloud. Once validated, these approaches can in principle dramatically reduce the density of ground check points needed to validate a data product. We have introduced a lidar sensor model and rigorous error propagation into our workflow for processing Geiger-mode lidar data generated by the 3DEO Vulcan and Sequoia airborne lidar systems. We report on validation of the predictive utility of this rigorous error propagation approach using airborne data and ground truth. We analyze predictive accuracy for both unobscured scenes and foliated (partially obscured) scenes.

Dale Fried, 3DEO, Inc.

David Kelley, 3DEO, Inc.

10:45 AM – 11:00 AM – Research methods for validation of lidar surfaces in complex terrain

The goal of the study is to understand aerial lidar’s effectiveness in generating high-quality elevation models across various landscapes. This study demonstrates the use of ground based terrestrial lidar data and uncrewed aerial systems (UAS)-based lidar data for evaluating and validating 3DEP lidar data over complex terrain. We focus on assessing the horizontal and vertical accuracy of aerial lidar data in diverse landscapes while developing innovative checkpoint targets and evaluating geometric consistency with existing lidar datasets. A primary objective is to assess lidar’s capability to map ground features beneath dense vegetation canopies. High-resolution terrestrial lidar scan data, collected in both open areas and under canopy, were used as a reference for validating the accuracy of the aerial lidar data across two case studies.

Case Study 1: Colorado (Urban and Complex Terrain) – This case encompasses an urban landscape alongside a region with unique rock outcroppings, sandstone hogbacks, open meadows, shrubby hillsides, and ponderosa pine forests. Validation methods included comparisons with high-resolution ground-based terrestrial lidar scans, high-accuracy ground control points (GCPs) established using GNSS surveys, and UAS-based lidar scans.

Case Study 2: Washington (Coastal and Diverse Forest) – The second case study focused on the dense forest canopies of northwestern Washington, which include various coastal and forested environments, as well as temperate rainforests. Validation methods involved high-resolution ground-based laser scan data, data from total stations, and the surveying of natural targets such as tree trunks.

The integration of terrestrial lidar scanning and total station surveys, along with low-flying UAS-based lidar data, as validation tools offers a promising approach to improving the accuracy and reliability of aerial lidar data, particularly in challenging environments.

Jeff Irwin, U.S. Geological Survey

Aparajithan Sampath, KBR Contractor to U.S. Geological Survey

Mark Bauer, U.S. Geological Survey

11:00 AM – 11:15 AM – Real-time geo-referencing in wide-area and corridor airborne laser scanning – technical challenges, accuracy limits, and emerging applications

Real-time geo-referencing in wide-area airborne laser scanning (ALS) has gained increasing interest in recent years, largely driven by the availability of GNSS auxiliary services which now enable point cloud accuracies that are sufficient for a variety of practical applications. This presentation explores the challenges involved in integrating state-of-the-art airborne laser scanners, GNSS/IMU systems, with real-time data processing workflows. A key issue is the substantial computational demand imposed by high pulse-rate scanners, which generate vast volumes of data requiring data pre-processing, range-ambiguity resolution, geo-referencing, data storage, and optional data streaming. The use of highly efficient software modules – similar to those employed in post-processing solutions – is a crucial strategy to ensure timely data fusion despite these high data rates. Special emphasis is placed on the accuracy limits achievable with currently available GNSS real-time aiding technologies. Finally, emerging applications are discussed, including real-time terrain modelling for, e.g., disaster management, corridor mapping for power line maintenance and damage assessment, and dynamic monitoring of infrastructure and construction sites. In this talk we provide a comprehensive overview of the current state-of-the-art and outline future directions for real-time geo-referencing in airborne laser scanning.

Peter Rieger, RIEGL Laser Measurement Systems GmbH

11:15 AM – 11:30 AM – Lidar Misconceptions & Truths: Debunking Common Data Assumptions

In the world of lidar, clean data doesn’t always mean correct data. A dataset may pass technical specifications and automated checks, yet still contain critical flaws that impact usability, interpretation, or decision-making. As a quality professional, we are often the last line of defense between seemingly high-quality deliverables and the end users who depend on that data for real-world applications.

This presentation will take a deep dive into common assumptions made about lidar data quality and how those assumptions can lead to overlooked issues. Lidar datasets that look visually complete may still contain inconsistencies in density, incorrect classifications, poor breakline integration, or artifacts in DEMs. Additionally, project specifications, while essential, can often leave room for ambiguity or misinterpretation across vendors, clients, and review teams.

From trusting automated checks too much to misinterpreting actual data errors, real examples will be shown to provide an understanding of what it means for a dataset to be “correct” or “compliant”.

Meagan Anderson, Dewberry

11:30 AM – 11:45 AM – Lidar data evaluation using cloud, open source data and software

The U.S. Geological Survey managed 3D Elevation Program (3DEP) has significantly expanded national lidar coverage, necessitating scalable, reproducible methods for assessing data quality across diverse terrains and acquisition conditions. This study builds on our cloud-native, open-source workflows designed to evaluate the geometric accuracy and consistency of 3DEP lidar datasets at a national scale. Leveraging tools such as PDAL, Open3D, and AWS infrastructure, the workflow integrates GNSS-surveyed ground control points and other data to validate airborne lidar collections. Three case studies demonstrate the application of this framework. In Arizona and parts of South Dakota, Nebraska and Iowa, the workflow confirmed the high vertical accuracy with minimal bias. In the second case study, we scale up the process and demonstrate the use of nationwide database of check points to assess the vertical accuracy of 3DEP data. In Colorado, we demonstrate the use of open-source building footprint databases (https://wiki.openstreetmap.org/wiki/Microsoft_Building_Footprint_Data) to aid large scale lidar data consistency evaluation across multiple time scales. The results underscore the effectiveness of combining cloud computing with open-source tools to perform large-scale lidar data quality assessments. This approach offers a reproducible, efficient solution for nationwide validation of 3DEP datasets, supporting enhanced decision-making in geospatial applications.

Aparajithan Sampath, KBR Contractor to U.S. Geological Survey

Jeff Irwin, U.S. Geological Survey

Barry Miller, U.S. Geological Survey

11:45 AM – 12:00 PM – Calibration and Validation of Inland Topobathymetric Lidar and Integrated Topobathy Models for Riverine Applications

Topobathymetric lidar has been widely applied to map river and shallow reservoir geomorphology and land-water elevations, advancing the knowledge needed to address environmental and ecological issues including protection of endangered habitats, flood hazards, landslides, management of rivers for recreational activities, and water supply analyses or modeling.  The U.S. Geological Survey (USGS) has been using topobathymetric lidar for mapping inland to coastal zone rivers and shallow reservoirs and establishing geospatial methods for the integration of inland surface and submerged elevation source datasets.  This paper describes outcomes from the 3D Elevation Program (3DEP) Allegheny River, PA topobathymetric lidar validation survey by presenting complete horizontal and vertical accuracy results and highlights the value of using sonar to independently validate bathymetric bottom classifications. A geospatial framework for integrating riverine and coastal topo-bathy elevation data into a common database aligned both vertically and horizontally to a common reference system are demonstrated for Southeast Texas and the Potomac River. Potomac River integrated data will benefit modeling for drinking water supply safety, flood prevention, and mitigating Atlantic sturgeon and American eel habitat losses.  Finally, a new innovative integration method will be demonstrated that uses enhanced binary stacking and spatial blending to improve seamless transition between disparate data sources.

Jeffrey Danielson, U.S. Geological Survey, Earth Resources Observation and Science (EROS) Center