August 24, 2009
3D Digital Elevation Models -- Accuracy and Use
Please note that contributed articles, blog entries, and comments posted on GIScafe.com are the views and opinion of the author and do not necessarily represent the views and opinions of the management and staff of Internet Business Systems and its subsidiary web-sites.
Welcome to GISWeekly!
GISWeekly examines select top news each week, picks out worthwhile reading from around the web, and special interest items you might not find elsewhere. This issue will feature Industry News, Top News of the Week, Acquisitions, Agreements, Alliances, Announcements, Financials, People, New Products, Around the Web and Events Calendar.
GISWeekly welcomes letters and feedback from readers, so let us know what you think. Send your comments to me at
Susan Smith, Managing Editor
3D Digital Elevation Models -- Accuracy and Use
By Susan Smith
In a recent webcast, Intermap Technologies, known for its NEXTMap 3D elevation models, talked about the importance of digital elevation models (DEM) and answered the question, “What is a DEM?”
Topics included: What different types of DEMS are there, different spatial data types associated with them, what technologies are used to create them, understanding vertical accuracy as well as posting cell size, and different types of applications that can benefit from DEMs.
Kevin Thomas, vice president, marketing and David Buhner, manager of data processing for Intermap explained the difference between vector based and raster data. Vector based data has points with an xy coordinate. It gives data a lat long grid coordinate to understand location. Points can be connected using lines, and lines between points give you distance between points. To take points and close them gives you a polygon. Vector based is for the most part just xy.
Raster data: there are two types of raster data, the first being imagery. Different types of technologies can collect imagery. With image data you can understand that if you get down to each pixel in an image, it can give you attribute such as color, or greyscale that will be a shade. The other type of raster data is DEM data, which lets you understand elevation with color.
Each “cell” in a 3D DEM contains an elevation value, according to Thomas. A DEM can be lain on its side to get a better visualization of what the terrain is like.
How are DEMs produced?
Photogrammetry was the first remote sensing technology developed, which involves a camera on the front of a plane and on the wings. They would use triangulation to understand the distance between plane and the ground, based on photography. With stereophotogrammetry there is a laser on the bottom of the plane, and the goal is to find precise elevation data between two points. LiDAR has laser pulsing and a point based distance system that can calculate distance between distance of land and elevation distance. There are GPS stations with both LiDAR and radar which collect survey points which assist in the surveying and processing of data.
The Shuttle Radar Topography Mission (SRTM) obtained elevation data on a near-global scale to generate the most complete high-resolution digital topographic database of earth. SRTM was comprised of a specially modified radar system that flew onboard the Space Shuttle Endeavour during an 11-day mission in February of 2000. It covered the majority of the earth and collected raw radar data of 9 terabytes, which was considered a huge amount of data at that time. SRTM is still used worldwide.
The means of collecting different types of DEM data varies. IFSAR has an accuracy of 1 meter. Photogrammetry and LiDAR have the most accurate data at 20 centimeter vertical accuracy.
Buchner said you need to look at what you’re trying to accomplish. If you’re doing a floodplain, 20 feet won’t be accurate enough. “You need to ask the right questions if you’re sourcing data,” he said.
LiDAR is the most accurate of the platforms, but very costly, as it takes long time to collect. Photogrammetry, is also some of the most accurate but time consuming. SPOT satellite and Intermap IFSAR are not quite as accurate but offer 1 meter data.
There are measurements that are predicted as opposed to those that are actual. Prices range according to vertical accuracy. “In the case of vertical accuracy, we take our dataset and compare it against a dataset that is deemed as truth and RMSE vertical accuracy,” said Buchner. “We don’t say our data is perfect for all applications. If you’re doing a small engineering project for a small area and you need highly accurate elevation data LiDAR is a great solution for you.”
Buchner added that for large areas like a metro area or state, this type of data might be overkill. One meter for a wider area is great as it saves costs, and can blend with 27 meter LiDAR for close up smaller areas as needed.
Cell size represents the spacing between your points, and different postings range from 30 meter all the way to 1 meter. This means that every 30 meters you’ll have a pixel or a point, and have an xy coordinate associated with it. With one meter you have more dense data. If you have postings of a smaller nature, the data will be more accurate.
A one-meter posted product for an entire metro area is a lot of data to deal with. Some people have to resample it down to a 25 meter product because it becomes too much data.
The horizontal accuracy of data has to do with the accuracy of points, but it is in conjunction with posting and vertical accuracy. With the size of your posting goes the size of your data set.
Types of DEMs available
A digital surface model (DSM) includes tower, trees, anything grown or built up on land. Whether radar or LiDAR – the view is what you would see from a birdseye view or higher.
Digital terrain models (DTMs) take out such features, manmade features, brush or trees to show true bare earth. Some radar can see through the earth’s crust, but most will bounce off what it first comes in contact with. “There are also ways and types of software and look at a DSM and see different items in surface, create algorithms and understand what’s underneath,” said Buchner.
Intermap has DSM and scraping of the earth in its model to make sure it’s removing the artifacts. “We have a team of employees in the production facility that manually go into every single tile and square mile and manually ensure the bald earth is as is before it goes to a validation group to ensure the product that is put on the shelf is correct. The customer doesn’t have to do anything, just plug it in to their application,” said Thomas.
Different applications for a DEM include: contour generation, enhanced mapping applications, viewshed analysis, flood mapping, telecommunications such as cell tower coverage, 3D road generation, geology, urban mapping, vegetation, wind power generation and watershed analysis.
Intermap’s high-resolution NEXTMap 3D DEMs were accepted as a source for updating FEMA flood maps this week as well, announced at the recent Association of State Floodplain Managers (ASFPM) national conference in Orlando, Florida.
The company’s NextMap DEM data has been used in Europe to update flood risk maps for the insurance industry, and flood risk products based on the NextMap data are used in UK, Germany and Switzerland. The addition to FEMA flood risk maps should make a large difference to the agency’s ability to anticipate and respond to flood crisis situations.
You can find the full GISCafe event calendar here.
To read more news, click here.
-- Susan Smith, GISCafe.com Managing Editor.
Be the first to review this article