A Detailed Primer on Digital Elevation Models
Avatar photo Nash Prado

A Detailed Primer on Digital Elevation Models

Digital Elevation Model | Infrastructure

In this article, you’ll learn about Digital Elevation Models, Digital Terrain Models, and Digital Surface Models. You’ll be able to understand how they are used and what industries benefit most.


Processed orthorectified 3D rendered satellite imagery captured by Kompsat-3 over Canberra Australia. The elevation data was derived from Tri-stereo Kompsat-3 image and processed by the Proimagery team


Digital Elevation Models (DEM) show a 3-dimensional view of a particular area of interest. DEMs provide valuable insights to engineers and GIS specialists, saving time and resources in analyzing large swaths of areas on foot.

Digital Elevation Models (DEM) are spatial data that represent the relief of a surface between points with known elevation. It represents the 3-dimensional values of a terrain’s surface. DEM data contains elevation values of the terrain over a specific area usually at a fixed grid of the bare earth. DEM has two types, the Digital Surface Model (DSM) and the Digital Terrain Model (DTM).


Differences of DSM and DTM

Digital Terrain Model is can be expressed as a three-dimensional representation of the earth’s surface that consists of X, Y, Z coordinates which represent latitude, longitude, and elevation respectively. Ground features such as mountains, rivers, and ridges can be determined from the Digital Terrain Model. A DTM is sometimes referred to as a DEM and also synonymous with Bare Earth

While Digital Surface Models represent the mean sea level elevations of the ground features from the surface of the earth such as trees, buildings, and other things above the Bare Earth. In short, the DSM shows features of the earth’s surface including all objects on it.



Satellite-derived elevation models

Proimagery can provide customers with an orthorectified very high-resolution single shot, stereo pair, and tri-stereo pair satellite imagery that can be processed for visualizing two-dimensional (2D) and three-dimensional (3D) terrain features of satellite imagery.

Stereo satellite imaging called stereoscopy or 3D imaging is a photogrammetry technique developed for creating a visualization of the depth of an image or sets of imagery. Satellite sensors can be tasked to collect either stereo imagery or tri-stereo imagery for the production of high-quality Digital Elevation Models.


Stereo pair imagery vs. tri-stereo pair images

Unlike the standard stereo pair satellite image which captures satellite images from just two angles, it cannot capture information in between buildings. But Tri-stereo acquisition captures images from three angles so that it can provide information that is hidden on steep terrains, clustered buildings, and densely built-up areas.

Proimagery can prepare Digital Elevation Models and Digital Surface Models using a variety of satellite sensors including:

This variety of satellite images can provide up-to-date digital elevation models and support project planners, emergency managers, logistics planning for field operations. All of this can be done without going into the field.


What are Digital Elevation Models used for?

The DEM industry has been constantly evolving since its inception. Improvements in optical and LiDAR sensor quality have resulted in increased resolution over time. When paired with a falling price per square kilometer of data received, you’ll see a trend that more and more use cases are now becoming economically viable. 20 years ago, most DEMs were used in Military and Agriculture applications. Now, you see many different use cases emerging. Here are just a few specific examples:

  • Volumetric Analysis – Commodities traders are using DEMs to quantify stockpiles of raw materials to see forecast future shortfalls & supply gluts
  • Insurance – Insurers can more accurately assess flood risk using real-world elevation data
  • Disaster Prevention – Forest fire & landslide risk can be remotely assessed, quantified and a mitigation strategy can be created before fieldwork is needed
  • Entertainment – Video games can use real-world 3D models of landscapes, rather than needing to generate an area from scratch saving time and money while adding realism to the gameplay
  • Telecommunications – Telecoms can accurately plan their 4G and 5G rollouts and estimate signal quality
  • Military and Defense – Governments can plan ideal areas to construct military installations
  • Energy – Home solar installers can accurately estimate solar radiation and generation potential based on roof angle for an entire neighborhood without needing to send a technician out.
  • And many more

Each use case has unique needs including specific requirements for horizontal resolution and vertical accuracy, or industry-standard data formats. In some cases, LiDAR is ideal; in others, photogrammetry works better. If you’re not sure what you need, Proimagery has a team of engineers who specialize in applying DEMs to different use cases.


LiDAR vs. Photogrammetry

LiDAR has been very popular in the news with significant advances in autonomous driving. But the reality is that LiDAR has been an industry-standard modeling technology for 30+ years. Photogrammetry has been used for even longer. These sensors both provide a 3D visualization of landscape, but in different ways.

LiDAR – Light Detection and Ranging is a remote sensing technology that sends many pulses of light towards a target, with a scanner to detect the data returned from each light pulse. The data returned can accurately estimate the geolocation of the returned data. When you model thousands or millions of returned data points, you create a point cloud of data like the building below.

Point Cloud from Autodesk

Image credit: Autodesk

Photogrammetry captures distance data in a way similar to your eyes. Images are taken from different angles around an object. When taking imagery of the same subject from different angles, distance can be effectively estimated even when the camera is very far away. Satellites flying 550 km above the earth can estimate the height of an object with a vertical accuracy of 2-3 meters.

Photogrammetry and LiDAR are different sensor technologies that get to the same result – A 3D model of a specific area of interest. It’s important to remember that while LiDAR may be superior in some use cases, photogrammetry is superior in others. In most cases, the sensor is far less important than price, timeliness, resolution, and accuracy.


Resolution & Accuracy

When choosing between data sources, data resolution and accuracy are usually the most important questions we get asked. The final use case will often determine the specific data source we use for a given project.

Horizontal Resolution (also referred to as horizontal spacing) – the level of detail in an image or model. The smaller the pixel size, the higher the level of data. Keep in mind that sometimes, more data is not ideal. Simulations for example often require many servers to compute results, even when using data from a lower resolution model.

Vertical Accuracy – the accuracy of the height value using a statistical model, calculated by comparing estimates from the DEM to known ground control points. The smaller the vertical accuracy number, the more accurate the data.


SensorDrone ImageryAirborne LiDARSkysat Imagery
Horizontal Resolution5 cm50 cm1.5 meters
Vertical Accuracy (RMSE)+/- 5 cm+/- 10 cm+/- 2 meters
All-in Price per km2$500$250$45 – $80


You’ll see from the table above that while Skysat (satellite) imagery is much cheaper than drone or Airborne LiDAR, it is also much lower in both resolution and accuracy.

The specific resolution you need will depend on several factors including how fast you require the data, the size of your area of interest, and the level of resolution & accuracy required. Reach out to Proimagery if you’d like to discuss your specific project.


Where can I find DEM data?

High-Resolution LiDAR DEM data is freely available provided by large-scale government projects. Here are the more popular LiDAR data sources we work with:

For areas not covered by high-resolution LiDAR, low-resolution 30-meter DEMs exist for the entire world

Proimagery can also prepare DEMs at any resolution for nearly any area using drone or satellite imagery. And if free data exists for your area of interest at the required level of accuracy, the engineers at Proimagery can help you turn that data into real-world insights.


What software can I use to open Digital Elevation Models?

If you’ve ordered or sourced a free DEM online that suits your area of interest, you’ll need special software to open the image. GeoTiff is the most common data format for DEMs, followed by .las / .laz for LiDAR point clouds.

QGIS is free and open-source software that can process GeoTiff DEMs and las point clouds.

Digital Elevation Model opened in QGIS software


ArcGIS Pro is enterprise-grade software that has built-in workflows that can leverage GeoTiff DEMs and las point clouds to analyze data in different ways

Digital Elevation Model opened in ArcGIS Pro

Image Credit: ESRI


Digital Surface Models, Digital Terrain Models / Digital Terrain models are 3D representations of the planet. They’re used in a multitude of industries for planning and analysis. Proimagery can source the necessary drone/satellite imagery and process Digital Elevation Models for interested clients. If you have a project in mind or want to know how you can leverage DEMs in your business, please reach out to us.