Statewide USGS 3DEP Lidar Topographic Differencing Applied to Indiana, USA

Abstract

Differencing multi-temporal topographic data (radar, lidar, or photogrammetrically derived point clouds or digital elevation models—DEMs) measures landscape change, with broad applications for scientific research, hazard management, industry, and urban planning. The United States Geological Survey’s 3D Elevation Program (3DEP) is an ambitious effort to collect light detection and ranging (lidar) topography over the United States’ lower 48 and Interferometric Synthetic Aperture Radar (IfSAR) in Alaska by 2023. The datasets collected through this program present an important opportunity to characterize topography and topographic change at regional and national scales. We present Indiana statewide topographic differencing results produced from the 2011–2013 and 2016–2020 lidar collections. We discuss the insights, challenges, and lessons learned from conducting large-scale differencing. Challenges include: (1) designing and implementing an automated differencing workflow over 94,000 km2 of high-resolution topography data, (2) ensuring sufficient computing resources, and (3) managing the analysis and visualization of the multiple terabytes of data. We highlight observations including infrastructure development, vegetation growth, and landscape change driven by agricultural practices, fluvial processes, and natural resource extraction. With 3DEP and the U.S. Interagency Elevation Inventory data, at least 37% of the Contiguous 48 U.S. states are already covered by repeat, openly available, high-resolution topography datasets, making topographic differencing possible.

Regional‐Scale Landscape Response to an Extreme Precipitation Event From Repeat Lidar and Object‐Based Image Analysis

“…Initial attempts to use repeat lidar scans to quantify changes in the elevation of the Earth's surface often reveal deficiencies and errors in the individual datasets or reveal mismatches in data characteristics (Cavalli et al., 2016; DeLong et al., 2012; Lallias‐Tacon et al., 2014; Schaffrath et al., 2015; Scott et al., 2022). This can be the case even when individual datasets meet the specifications given to the vendors, and after the data pass quality assurance and control checks.…”

“…When comparing two lidar datasets, there are many possible causes of apparent error (Passalacqua et al., 2015; Scott et al., 2022). Processing airborne lidar requires a complex workflow that involves data from various sensors (i.e., laser ranging, inertial measurement unit, global navigation satellite system sensors on the aircraft and on the ground) that produce disparate measurement data that must be assimilated to generate raw point‐cloud data.…”

“…These challenges may include the need to improve comparisons of multi‐temporal data hampered by positioning error, internal flightline misalignment (we refer to the swath of point data collected on each pass by a survey aircraft as a “flightline”, and misalignment refers to the difference in position of the overlapping portion of point data from overlapping flightlines), sensor calibration error, challenging coordinate system and unit transformations, effects of vegetation, contrasts in measurement density, differences in how point clouds are classified with respect to bare ground, vegetation, infrastructure and water bodies, and challenges related to the subsequent analyses and interpretation of change detection results (e.g., Bernard et al., 2021; Bull et al., 2010; Burns et al., 2010; Daehne & Corsini, 2013; Glennie et al., 2014). These challenges are manifold at larger spatial scales (Scott et al., 2022), with data of variable age and provenance, when studying a range of surface processes across a diversity of landforms, and when vegetation cover changes along with the geomorphic change. Furthermore, the results from change detection over larger spatial scales, even if performed using two‐dimensional grids of elevation rather than full three‐dimensional analyses, may be challenging to analyze, classify and interpret because of their high resolution and the high variability in pixel values over short distances.…”

“…Although lidar and other high‐resolution elevation datasets generally do an excellent job at characterizing landscapes in three dimensions at a scale sufficient to interpret the nature of active surface processes, when data collected at different times are compared, limitations and error sources become more apparent. As such, distilling geomorphic change from error and noise can prove challenging (Bernard et al., 2021; Burns et al., 2010; Glennie et al., 2014; Passalacqua et al., 2015; Schaffrath et al., 2015; Scott et al., 2022). Site‐specific change detection from repeat airborne lidar may now be considered routine, but, even at small spatial scales, unique challenges may require some amount of data reprocessing to optimize the recognition of real geomorphic change and minimize error.…”

U-net-based semantic classification for flood extent extraction using SAR imagery and GEE platform: A case study for 2019 central US flooding

“…Although the data source for the two DEM layers is not the same, we believe the performance difference is primarily due to the finer data resolution. This is because they are both obtained from US federal agencies and have been widely used in previous studies (Oh et al, 2022;Scott et al, 2022;Uuemaa et al, 2020;Wong et al, 2020), and thus even if there are slight vertical differences between these two products, it is less likely that one has significantly higher data accuracy than the other. We discovered that the 10m three-band model without pretrain outperformed the outcomes of the five-band 10m model without pretrain.…”