Earth Sci. Improving the TanDEM-X digital elevation model for flood modelling using flood extents from synthetic aperture radar images. McDougall, K. Mukherjee, S. Evaluation of vertical accuracy of open source digital elevation model DEM. Earth Obs. NIH Human Genome Project. Fact Sheet, National Institutes of Health. O'Loughlin, F. Patel, A. Space Sci. Rabus, B. The shuttle radar topography mission—a new class of digital elevation models acquired by spaceborne radar.

Riegler, G. Robinson, N. Sampson, C. A high-resolution global flood hazard model. Perspectives on open access high resolution digital elevation models to produce global flood hazard layers. Sanders, B. Evaluation of on-line DEMs for flood inundation modeling. Schumann, G. Near real-time flood wave approximation on large rivers from space: application to the River Po, Italy.

Fight floods on a global scale. Nature High-accuracy elevation data at large scales from airborne single-pass SAR interferometry. Vricon The Globe in 3D. PubMed Abstract. Walker, J. On the effect of digital elevation model accuracy on hydrology and geomorphology.

Yamazaki, D. A high-accuracy map of global terrain elevations. Yue, L. Elementary Surveying: An Introduction to Geomatics. Pren-tice Hall, New Jersey, The mean higher high water MHHW line is one of the most widely-used tide datum parameters in risk assessment, being used as the extreme vertical reference in the studies of Murdukhayeva et al.

This can become a problem for coastal management and civil defense based on risk charts, given that it is important to use flood thresholds based on the tide datum WDOWINSKI et al.

Schimid et al. This may represent an important source of error, depending on the reference used for the adjustment of the geoid model, especially if the study area has a large tidal amplitude. Concerns over the lack of vertical consistency among the different data used in coastal research in the United States led to the National Oceanic and Atmospheric Administration NOAA producing an adjustment tool called VDatum.

This software was designed to convert geospatial data among a range of altimetric references used in the United States, whether derived from tides, or orthometric geoid or geometric ellipsoidal sources. One other limitation associated with the tide datum is the accessibility of the tide-measuring instruments. Most ports and waterways use tide gauges, but the availability of the data depends on the distribution network SILVA et al.

Many of these tide gauges are also installed in sheltered locations that are influenced intensely by rivers, such as deltas and estuaries, making inferences impossible for adjacent areas that are exposed directly to meteoceanographic forces, as shown by Goulart GOULART, E. Dissertation, Universidade Federal do Rio Grande, Although a tide table based on astronomical parameters provides the predicted tide at a given latitude, it does not substitute the historical records from a tide gauge, which will normally be linked directly to the reference body of water.

However, in the absence of local tide data or where the tide datum is incompatible, there are other ways of estimating the vertical reference level of the coastline. Shoreline Definition and Detection: A review. Journal of Coastal Research 21 4 , The principal limitations of this method are related to its reduced temporal representativeness, which restricts the sample to the morphodynamic behavior of the beach system for any given period.

Topographic surveys, regardless of the data collection mode, may represent only the seasonal or daily characteristics of the local sea level due to the high level of variability of the transport rates and the typical sedimentation patterns of the beach environment Figure 5. It is possible to estimate the reference tide level of a beach using a series of video images.

This permits the horizontal position of the water line to be estimated using digital image processing algorithms. The reduction in the level of the ellipsoidal datum through the calculation of the geoid undulation is another technique used frequently in flood analysis WEBSTER et al.

Using topographic lidar to map flood risk from storm-surge events for Charlottetown, Prince Edward Island, Canada. Canadian Journal Remote Sensing 30 1 , Ellipsoidal systems consider the Earth's surface to be a geometrically perfect ellipsoid with constant gravimetric potential and rotation around its polar axis, whereas geodetic systems represent the terrestrial surface with its irregularities in the form of non-uniform heights.

The irregularity of the geoid representation is due to the variation in the distribution of the density and mass of the planet, in addition to its rotation, which results in a non-homogeneous distribution of the terrestrial gravitational field. Using the gravimetric data in the form of a geoidal undulation N , it is possible to determine the height difference between the ellipsoidal data collection system and the geodetic reference system, thus obtaining the orthometric height H from equation 1, following Monico MONICO, J.

The vertical adjustment provided by the geoid undulation is fundamental to the adoption of orthometric heights in coastal research, especially in studies that focus on the hydrological dynamics of the continent-ocean interface. Concerns on this imprecision, as highlighted above, are justified, given that, in many cases, the vertical error may exceed the sea elevation predicted by projected scenarios of sea-level rise GESCH, GESCH, D.

The bathtub model LEON et al,. The bathtub approach is used primarily for the assessment of the flooding potential of coastal areas flood inundation vulnerability or is associated with demographic data and infor-mation on infrastructure to determine the flood inundation risk. In the modern conception of this approach, it is assumed that the body of water will in-clude all the land located at altitudes below the projected water level, given that there is a direct connection with the source of the flood or with the flooded cells.

The bathtub approach is wide-ly used in the assessment models of climate change impacts related to sea-level rise SEYATH et al. Despite its ample use, the bathtub approach has a number of limitations and demands certain precautions for its application. Both Schimid et al. Effects of sea-level rise on barrier island groundwater system dynamics - ecohydrological implications. Ecohydrology 7, Figure 6. The simple bathtub modeling of flooding is generally used in low-resolution digital sur-face models, which implies a series of restrictions for the analysis of coastal flooding YUNUS et al.

In recent years, however, a number of authors have introduced a more complex ap-proach to the flood models based on the bathtub approach, which makes them more versatile. The first of these aspects refers to the adjacent displacement or, in hydro-logical terms, the insertion of the surface flow, which is associated with the scale of the data the detail of the morphological features and the spreading rule cell connectivity and runoff coeffi-cient adopted in the study, which had rarely been employed in coastal flood models, but were widespread in studies of the drainage systems of hydrographic basins.

According to Longley et al. The water displace-ment function is commonly known as a spread, and consists of the total friction calculated for each of the possible paths established by the rules of displacement LONGLEY et al. Essentially, a displacement rule is selected for the hydrological model in which the flow priori-tizes the path with the least friction, from a given set of possible paths. The rules used most fre-quently include the zero-way, the four-way and the eight-way.

The zero-way rule, which was presented above, is applied in the simple bathtub ap-proach, in which there is no hydrological connectivity among the cells no displacement.

This single condition rule states that the cell will be flooded instantaneously if its elevation is lower than the projected sea level YUNUS et al. By contrast, the four-way and eight-way rules establish paths connecting adjacent cells Figure 7. The four-way rule is based on the con-nection of the cells located in the four cardinal positions, evaluating paths in four possible direc-tions.

The eight-way rule, adds the four diagonal axes, permitting the evaluation of eight possi-ble path directions. As Longley et al. These models predict coastal flooding if two conditions are met:. In the latter case, the water may flow into any of the neighboring cells, according to the displacement rule four-way or eight-way moving in the direction of the lowest friction, accord-ing to the slope YUNUS et al.

These authors pointed out that, while the four-way rule may underestimate the flow connections because it presents only four possible paths, the introduction of the diagonal paths may overestimate connectivity in the eight-way rule. In both cases, however, the higher connectivity tends to enrich the micro-features of the relief obtained by high-resolution DEM.

Yunus et al. The DEM usually refers to a digital representation of the earth's surface, but if it contains data on the height of targets that are above the ground, it is considered to be a Digital Surface Model DSM , while the Digital Terrain Mod-el DTM is a surface model that includes only the ground elevation, with minimal interference from other objects Figure 8.

Paradoxically, the high resolution of the DEMs, provided by the modern tools of topo-graphic data acquisition and processing, bring new concerns with regard to the hydrological connectivity in the flood models.

In coastal flooding models, the data associated with features of high verticality may cover depres-sions and natural drainage channels which provide connectivity, but may not be represented in the surface model.

In urban areas, infrastructure such as buildings, bridges, and artificial drainage systems will be georeferenced from the ground elevation, to which the respective height is added. Rural areas are not different, given that tall reference points such as trees and transmission towers have the same limitation for the definition of the elevation of features that are above ground level. It is important to note that this effect is due not only to the verti-cal increment promoted by the features, but also to the blocking of channels and the masking of the depressions located underneath them.

Furthermore, this and other metrics internal smoothing dR , plane-fit RMSE , provide an additional visual assessment of DEM quality and the spatial patterns of noise, closely tied to hillshade observations Figure 2. B 2D DFT periodogram, showing the normalized amplitude power in frequency space.

As in Figure 4 , for visualization purposes the 2D DFT was morphologically dilated to increase the size of the discrete, high power peaks and values below the 50th percentile were excluded colored white. Without a local detrending procedure to accentuate inter-pixel consistency e.

In this example, the approximately north-south trending ridges create the highest power signals in B. In another step, we tested the result of increasing the kernel size for high-pass filtering in the HPHS calculation Supplementary Figure S1. A recent review by Polidori and El Hage highlights the need for different approaches of inter-pixel consistency reporting to improve understanding of DEM quality beyond vertical accuracy. This is key given increases in quantitative geomorphometry and dissemination of DEMs from many sometimes poorly understood sources Sofia, The metrics developed here provide a more suitable assessment of DEM quality compared to point-based vertical accuracy, which does not account for the spatial variability of DEM vertical errors that impact derivatives of elevation.

These metrics do not use reference surfaces e. Here, we build on this previous work and develop another approach to quantify the wavelengths and orientations of these artifacts. The mast-oscillations during collection are responsible for the original artifacts Farr et al.

These artifacts could be removed via band-pass filtering on the selected wavelengths identified as spectral peaks Yamazaki et al. All five DEMs are delivered with auxiliary rasters containing pixel-level information on their source. Increasing the number of stacked scenes, especially in complex topography, can improve DEM quality e.

In any case, these void- and stack-masks, along with other auxiliary files, can be a useful check on DEM quality. Our goal here was to consider only the elevation surface to extract internal error metrics without any other reference data or any background DEM processing data, which is how many end users receive and utilize the gridded DEMs.

We do note that observations in the study area show a smoother surface in the Copernicus DEM compared with TanDEM-X; however, in local areas this smoothing has noticeably flattened true topographic expression in the form of rough bedrock outcrops. This is an inevitable result of automatic DEM editing, which often has significant moving-window smoothing steps. Other recent efforts towards DEM editing and fusion e.

Any DEM created by multi-step editing of spaceborne data should be treated with care and analyzed for inter-pixel consistency, in addition to traditional vertical accuracy metrics. High power in adjacent pixel steps measured on the HPHS grid Figure 8 are a sign of low inter-pixel consistency in our geomorphically smooth, nearly vegetation-free study area.

The high-power in adjacent pixels can be accounted for by sensor errors signal to noise ratio and speckle associated with radar interferometric generation Buckley et al. This again highlights the benefit of the HPHS calculation and Fourier analysis to identify resampling artifacts, which may vary for different resampling schemes. Resampling during reprojection is a common pre-processing step prior to any topographic analysis and is a requirement by such software as TopoToolbox Schwanghart and Scherler, and LSDTopoTools Mudd et al.

Thus, the differences in high-frequency adjacent pixel variance for different resampling schemes are particularly notable results of the analysis Figure 8 ; Supplementary Figure S2. Often resampling is done using bilinear or nearest-neighbor approaches, as these are quick and thought to provide reasonable results.

Resampling methods using larger or adjustable window sizes or higher-order polynomials, such as cubic spline resampling reduces the variance in adjacent pixels and increases inter-pixel consistency. However, cubic spline resampling may also be over-smoothing locations with high topographic variability at short distances. This is a natural result of DEM smoothing and points to a nuance of m DEM usage: the topographic variability at short distances is convolved with the signal of non-topographic variance.

Smoothing these DEMs to increase inter-pixel consistency while retaining topographic signatures will require adaptive resampling schemes. In any case, going forward with the geomorphic analysis we use the cubic spline resampling. Importantly, we note that different resampling schemes only change the magnitude and not pattern relative difference between DEMs of the results.

Moving from a tile-based approach to quantify the variance in each DEM at specific wavelengths pixel steps , we turn to our selected catchments Figure 1 to assess the impact of inter-pixel consistency differences for geomorphic research.

The slope distributions calculated for each catchment and each 30 m cubic spline resampled DEM using the Zevenbergen and Thorne algorithm are presented in Figure The density was calculated using a kernel-density estimate of the underlying distribution. Slope was calculated with the Zevenbergen and Thorne algorithm on cubic spline resampled m DEMs but results are comparable with other resampling methods.

To further investigate the impact on the slope distribution, we use percentile-percentile plots a. In this case, we combine the measurements for all three catchments, as the individual catchment plots showed similar relationships Supplementary Figures S3—S5 , and Figure 11 presents an average of this. We note that the Zevenbergen and Thorne algorithm takes the gradient from adjacent edges touching pixels for slope calculations.

In the Supplementary Figure S6 , we also use the Horn algorithm, which considers the diagonally adjacent corners touching pixels, and may be more appropriate for rougher surfaces.

In Supplementary Figure S6 , we note that the alternative slope calculation only reduces the magnitude of measured slopes e. Catchment slope percentile-percentile plots on cubic spline resampled m DEMs. All three slope distributions for the three test catchments Figure 1 are combined, with separate plots for each catchment shown in Supplementary Figures S3—S5.

The maximum relative percentage difference in slope compared to Copernicus approaches 0. Slope was calculated with the Zevenbergen and Thorne algorithm, and Supplementary Figure S6 presents the results using the alternative Horn algorithm. Therefore, studies relying on hillslope distributions from the ASTER and SRTM DEMs will likely under-estimate the central distribution median and over-estimate the tail steepest topography , which may impact conclusions of hillslope responses to changes in erosion e.

The inter-pixel consistency investigation is extended to the channel network in Figure The k sn calculation in other catchments is generally consistent, although we do note differences for more subtle, lower-magnitude possible knickpoints in the downstream Palermo profile Supplementary Figure S Comparison of trunk stream longitudinal river profile for the Honda Catchment Figure 1.

The middle row D—F shows the standard deviation of the gradient in 1-km flow distance bins, where the first row in these sub-plots correspond to the window used in A—C. The channel steepness profile analysis is a metric usually applied to length scales of several hundred meters or longer and thus performs inherent smoothing of the input DEM data.

This mitigates some effects of DEMs with large inter-pixel inconsistencies e. Therefore, the consistent performance of the k sn metric is expected, and our previous work Purinton and Bookhagen, showed similar concavity measurements using a similar range of DEMs and DEM resolutions. We argue that for river-profile steepness analysis over several-km flow lengths in steep mountains, where the rivers descend hundreds to thousands of m in elevation, the tested DEMs perform similarly.

However, we note that higher-resolution DEMs e. This is important for detection of high magnitude, but short length-scale slope changes in river profiles, for example for knickpoint and step-pool detection. The similarity in k sn is not matched by the patterns of local channel node to channel node gradient shown in Figure The differences in longitudinal river profile gradient calculations between the different DEMs are summarized in Figure This combines all data from the three catchment trunk streams and DEMs similar to Figure This implies that a DEM with lower inter-pixel consistency should use different larger window sizes for channel gradient calculations than a more internally consistent DEM.

Summary window size versus gradient standard deviation for all trunk stream river profiles. A Sum of the gradient standard deviations across all 1-km flow distance bins in each catchment for each DEM.

B Relationship between standard deviation of gradient and average channel gradient in each 1-km flow distance bin for all three catchments for ASTER-GDEMv3, with gradient calculated in an node window.

Higher gradients have higher standard deviation, suggesting that higher gradients have lower inter-pixel consistencies. Figure 13B shows another subtlety of the channel gradient analysis.

Here, we see that the variability in channel gradient increases with increasing average gradient, which means that steeper parts of the channel are expected to have a greater spread of steepness. This is exactly the nature of a concave channel profile, where gradient decreases quasi-exponentially downstream the values change more in the upper, steeper reaches. Thus, the window size of slope calculation will impact different channel reaches differently, which may be an important consideration over long or very steep profiles.

Recent work has explored the usage of adaptive approaches with smoothing adjusted to the amount of slope variability Schwanghart and Scherler, ; Gailleton et al. This is a possible sign of a hydrologic preconditioning step of this dataset, though unreported in the technical documentation EORC, Although longitudinal river profile analysis at large several km scales in steep, high-relief catchments for tectonic and climatic forcings are less affected by DEM choice e.

Furthermore, fine-scale analysis of channel gradients will be particularly impacted by the channel-node to channel-node variability, with implications for assessing reach-scale channel morphology and knickpoint detection for tectonic geomorphology e.

This study does not address DEMs of different resolution, as we focus only on the widely used and open-source near global 30 m DEMs. We do note that DEM resolution will impact geomorphic analysis and the calculation of derivatives of elevation Purinton and Bookhagen, ; Grieve et al.

Although the spatial resolution of these DEMs is too coarse for channel-head detection e. This is an important component in soil mantled and diffusional landscapes, irrespective of vegetation cover.

Previous work to quantify vertical error in DEMs e. In our study, a primary goal is to avoid reference data and develop an internal metric to compare DEM quality using only a priori knowledge of a smooth, bare-earth study area with mixed steep and flat terrain. While the methods developed by Becek demonstrate a novel use of airplane runways as continuous reference surfaces, this is limited to flat terrain of limited spatial extent and requires some assumptions, such as a uniform distribution of quantization and slope-induce errors.

Our local pixel variability metric based directly on the hillshade image which includes slope and aspect information , combined with quantification of inter-pixel consistency and demonstration of the impacts on geomorphic research, provides a quantitative explanation of spaceborne DEM quality.

We do note that other authors have found better performance of the ALOS-W3D for stream profile analysis Schwanghart and Scherler, , possibly due to hydrologic preconditioning. Our study area in the steep Central Andes is geomorphically smooth and arid nearly vegetation-free , which results in ideal conditions for bare-earth remote sensing and comparison of DEMs. We refer to this as the inter-pixel consistency, where low high inter-pixel consistency refers to high low variability in adjacent pixels.

Our chosen analysis metric is based on high-pass filtering of the hillshade images for each DEM and does not rely on reference data. We found:. These DEMs also contain significant high-frequency variability in adjacent pixel steps, detrimental to pixel neighborhood calculations, such as hillslope angle and channel gradient.

The caveat of larger window sizes or different resampling schemes is that true topographic variability is likely smoothed along with artifact variability. Therefore, the selection of a DEM with the most consistent height representation i. A more complete picture of DEM quality for geomorphic analysis includes pixel variability quantified with respect to local neighborhoods, beyond point-based vertical accuracy from reference data. Publicly available datasets were analyzed in this study.

BB and BP defined the project. BP carried out the analysis and lead the manuscript writing with input from BB. BB provided funding. The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher. Stefanie Tofelde, Taylor Smith, Ariane Mueting, Aljoscha Rheinwalt, and the rest of the Geological Remote Sensing group at the University of Potsdam are thanked for conversations and suggestions, particularly with regards to terminology.

Abrams, M. Google Scholar. Remote Sensing 12, Allmendinger, R. Earth Planet. Alonso, R. Giant Evaporite Belts of the Neogene central andes. Arrell, K. Earth Surf. Landforms 33, — Aster, G. Baade, J. Mapping, Remote Sensing, and Geospatial Data. Apply Filter. What is the difference between lidar data and a digital elevation model DEM?

Light detection and ranging lidar data are collected from aircraft using sensors that detect the reflections of a pulsed laser beam.

We will add the reported information to our "Digital Elevation Model Issues" data log for further analysis and possible correction. Spikes, pits, seam-line anomalies, and other data errors are of concern to us. We are working to correct or These interpolated point elevations are not official and do not represent precisely measured ground surveyed values. Elevations derived for a specific location using the Elevation Point Filter Total Items: 6. Year Published: 3D Elevation Program—Federal best practices The goal of the 3D Elevation Program 3DEP is to complete nationwide data acquisition in 8 years, by , to provide the first-ever national baseline of consistent high-resolution three-dimensional data—including bare earth elevations and three-dimensional point clouds—collected in a timeframe of less than a decade.

View Citation. Lukas, V. Geological Survey Fact Sheet —, 2 p. Year Published: The National Map—New data delivery homepage, advanced viewer, lidar visualization As one of the cornerstones of the U. Attribution: National Geospatial Program. Year Published: The National Map seamless digital elevation model specifications This specification documents the requirements and standards used to produce the seamless elevation layers for The National Map of the United States.

Archuleta, Christy-Ann M. Archuleta, C.