A Digital Terrain Model (DTM) represents the bare-earth surface of the Earth, showing elevation data with all natural and man-made features removed. Unlike a Digital Surface Model (DSM) which includes buildings and vegetation, a DTM reveals the underlying ground—essential for engineering, hydrology, and terrain analysis applications.

Think of a DTM as what the landscape would look like if you removed every building, tree, vehicle, and structure. Only the natural ground surface remains, showing valleys, ridges, slopes, and the true shape of the terrain.

Key Characteristics of Digital Terrain Models

Understanding DTM properties helps you select the right elevation product:

• Bare-earth only: All above-ground objects (buildings, vegetation, vehicles) are removed
• Ground-classified points: In LiDAR processing, DTMs are generated from points classified as ground
• Raster format: Typically stored as GeoTIFF with each pixel containing ground elevation
• May include breaklines: Vector features representing ridges, valleys, and drainage lines for improved accuracy
• Referenced to datum: Elevations relative to a vertical datum (WGS84, NAVD88, local geoid)

DTM vs DSM vs DEM: Understanding the Differences

These three terms are frequently confused. Here’s how they differ:

Digital Terrain Model (DTM)

What it captures: Bare ground only—vegetation, buildings, and all surface objects removed.

Best for: Hydrology, flood modeling, civil engineering, contour generation, slope analysis, earthworks planning.

Digital Surface Model (DSM)

What it captures: All visible surfaces including buildings, trees, and terrain.

Best for: Urban planning, telecommunications, forestry canopy analysis, obstacle detection.

Learn more in our Digital Surface Model guide.

Digital Elevation Model (DEM)

What it captures: Generic term for any gridded elevation dataset. Usage varies by region—in the US, DEM often means the same as DTM.

Quick Comparison Table

Feature	DTM	DSM	DEM
Shows bare ground	Yes	No	Varies
Includes buildings	No	Yes	Varies
Includes vegetation	No	Yes	Varies
Used for flood modeling	Yes	Rarely	Yes
Used for contours	Yes	No	Yes
Used for building height	No	Yes	No

How Digital Terrain Models Are Created

Creating an accurate DTM requires separating ground from non-ground features—a process called classification or filtering.

From LiDAR Point Clouds

LiDAR provides the most accurate method for DTM creation. The process involves:

1. Data acquisition: Airborne or drone LiDAR captures millions of elevation points
2. Point cloud classification: Algorithms identify ground points vs. vegetation, buildings, noise
3. Ground point extraction: Only ground-classified points used for DTM
4. Surface interpolation: Points converted to continuous raster surface
5. Export: DTM saved as GeoTIFF or other elevation format

LiDAR’s key advantage: laser pulses penetrate vegetation canopy, reaching the ground underneath trees where photogrammetry cannot.

From Photogrammetry

Photogrammetric DTMs require additional processing to remove vegetation:

• Start with DSM from stereo imagery
• Apply filtering algorithms to identify and remove above-ground objects
• Results are less accurate in vegetated areas since cameras cannot see through trees
• Best suited for open terrain or urban areas with visible ground

Classification Accuracy Matters

DTM quality depends entirely on accurate ground classification. Common challenges include:

• Dense vegetation: Limited ground returns under thick canopy
• Complex terrain: Steep slopes can confuse classification algorithms
• Low structures: Cars, fences, and low walls may be misclassified as ground
• Water surfaces: LiDAR struggles with specular water reflections

Lidarvisor’s automatic classification handles these challenges, correctly identifying ground points in diverse terrain conditions.

Digital Terrain Model Applications

DTMs serve critical roles wherever bare-earth elevation matters:

Hydrology and Flood Modeling

Water flows across terrain, not over rooftops. DTMs are essential for:

• Watershed delineation and drainage basin analysis
• Flow direction and accumulation modeling
• Flood extent prediction and risk mapping
• Stream network extraction
• Stormwater management design

Civil Engineering and Construction

Engineers need accurate ground surface data for:

• Cut/fill volume calculations for earthworks
• Road and railway corridor design
• Foundation planning and grading
• Drainage design and pipe slope calculations
• Contour line generation for site plans

Learn more about LiDAR for construction applications.

Land Surveying

Surveyors use DTMs for:

• Topographic surveys and mapping
• Boundary delineation based on terrain features
• Volume surveys for stockpiles and excavations
• CAD-ready contour deliverables via DXF export

Agriculture and Land Management

Agricultural applications include:

• Precision agriculture and variable rate application
• Drainage planning and tile layout
• Irrigation system design
• Erosion risk assessment
• Field boundary and terrace mapping

Forestry

While DSMs capture canopy, DTMs enable forest analysis by providing:

• Ground reference for Canopy Height Model (CHM = DSM – DTM)
• Terrain for harvest planning and road design
• Slope analysis for equipment access
• Erosion and stability assessment

Archaeology

DTMs reveal hidden features under vegetation:

• Ancient structures invisible under forest cover
• Subtle terrain features indicating buried remains
• Landscape-scale patterns of human activity
• Hillshade visualization for feature detection

Learn more about LiDAR for archaeology.

DTM Resolution and Accuracy

Choosing appropriate DTM resolution depends on your application:

Resolution Guidelines

Application	Recommended Resolution	Rationale
Regional flood modeling	5-10m	Captures major drainage patterns
Site engineering	0.5-1m	Resolves grading details
Detailed construction	0.1-0.25m	Captures fine grade breaks
Archaeological survey	0.25-0.5m	Reveals subtle features

Vertical Accuracy

DTM accuracy depends on source data quality:

• Airborne LiDAR: 5-15 cm RMSE vertical accuracy
• Drone LiDAR with GCPs: 3-10 cm RMSE
• Photogrammetry (open terrain): 10-50 cm RMSE
• Satellite-derived: 1-5 m RMSE

DTM Derivative Products

DTMs enable generation of multiple derivative products:

Contour Lines

Extract elevation contours at specified intervals for topographic mapping and CAD deliverables. Learn about contour generation.

Slope Maps

Calculate terrain gradient for stability analysis, equipment access planning, and erosion assessment.

Aspect Maps

Determine slope direction for solar exposure analysis, microclimate modeling, and habitat assessment.

Hillshade

Simulate illumination for terrain visualization, revealing subtle features invisible in raw elevation data.

Canopy Height Model (CHM)

Subtract DTM from DSM to calculate vegetation heights: CHM = DSM – DTM. Essential for forestry and canopy height analysis.

Creating DTMs with Lidarvisor

Lidarvisor automates DTM generation from LiDAR point clouds:

1. Upload: Submit your LAS or LAZ point cloud
2. Automatic classification: Ground points identified and separated from vegetation, buildings, vehicles
3. DTM generation: Bare-earth terrain model created from ground-classified points
4. Derivative products: Hillshade, contours, and slope maps generated automatically
5. Download: Export as GeoTIFF, DXF contours, or classified point cloud

No GIS expertise required. Upload your data and receive survey-ready terrain products in minutes.

Frequently Asked Questions

What is the difference between DTM and DEM?

DTM specifically means bare-earth terrain with all objects removed. DEM is a generic term for any elevation grid. In practice, the terms are often used interchangeably, especially in the United States where DEM typically refers to bare-earth data.

Why use DTM instead of DSM?

Use DTM when you need the actual ground surface—for hydrology, engineering, or any application where buildings and trees would interfere with analysis. Use DSM when above-ground features matter (urban planning, telecommunications, forestry canopy).

Can DTM be created from photogrammetry?

Yes, but with limitations. Photogrammetry produces a DSM first, then filtering algorithms attempt to remove vegetation. Results are less accurate under tree canopy since cameras cannot see through vegetation like LiDAR can.

What file formats are DTMs stored in?

DTMs are typically stored as GeoTIFF—georeferenced raster images where each pixel contains a ground elevation value. Other formats include ASCII grid, ESRI Grid, and various proprietary formats. Lidarvisor exports DTMs as GeoTIFF compatible with all major GIS software.

How accurate is a LiDAR-derived DTM?

LiDAR DTMs typically achieve 5-15 cm vertical accuracy depending on sensor quality, flight parameters, and ground conditions. Drone LiDAR with ground control points can achieve 3-10 cm accuracy. Classification quality affects final DTM accuracy—poorly classified data produces less accurate terrain models.

Conclusion

Digital Terrain Models provide the bare-earth foundation for countless applications—from flood modeling to construction planning to archaeological discovery. Understanding when to use DTM versus DSM ensures you select the right elevation product for your needs.

LiDAR technology delivers the most accurate DTMs by penetrating vegetation to capture true ground elevations. Modern processing platforms like Lidarvisor simplify the workflow, automatically classifying point clouds and generating terrain products ready for analysis.

Ready to create a DTM from your LiDAR data? Create a free Lidarvisor account and upload your point cloud. You’ll receive DTM, DSM, hillshade, and contour products—all from a single upload.

Free DTM Data Sources

Before creating your own DTM, check if free public data meets your needs. Several government agencies provide high-quality DTM data at no cost:

USGS 3DEP (United States)

The USGS 3D Elevation Program offers the most comprehensive elevation data coverage for the United States. Quality Level 2 (QL2) data provides 1-meter resolution with 10cm vertical accuracy—suitable for most engineering and planning applications. Access data through The National Map.

EU Copernicus DEM (Global)

The Copernicus DEM provides 30-meter and 90-meter global coverage derived from TanDEM-X radar data. While lower resolution than LiDAR-derived DTMs, it’s suitable for regional analysis, watershed studies, and areas without better alternatives. Available through the Copernicus Space Data Ecosystem.

SRTM (Global)

The Shuttle Radar Topography Mission provides 30-meter (1 arc-second) data for most of the world. Note that SRTM is technically a DSM, not a DTM—it includes vegetation and building heights. Best used where no bare-earth alternatives exist.

National and Regional Programs

• UK Environment Agency: 1m DTM covering England
• Australia Elvis: Various resolution DEMs across Australia
• OpenTopography: Curated LiDAR datasets from research projects worldwide

When free data doesn’t meet your accuracy or coverage requirements, creating a custom DTM from drone or aerial LiDAR provides the best results.

DTM File Formats Explained

DTMs can be stored in various formats, each with trade-offs for file size, precision, and software compatibility:

GeoTIFF (.tif)

The most widely used DTM format. GeoTIFF embeds coordinate reference system (CRS) information directly in the file, making it self-describing and compatible with virtually all GIS software. Supports 32-bit floating point precision for accurate elevation values. Recommended for most applications.

ASCII Grid (.asc)

A simple text-based format readable by humans and easily parsed by scripts. Header defines grid parameters (cell size, origin, no-data value), followed by elevation values in rows. Larger file sizes than binary formats but useful for data exchange and debugging.

ESRI Grid

Native raster format for ArcGIS products. Stored as a folder containing multiple files (binary data, header, statistics). Best suited for workflows entirely within the Esri ecosystem.

LAS/LAZ Point Clouds

While not a DTM format per se, classified LAS/LAZ files contain the ground points from which DTMs are generated. Lidarvisor can export ground-classified point clouds for users who need maximum flexibility in downstream processing.

DTM Interpolation Methods

When converting discrete elevation points into a continuous DTM surface, the interpolation method significantly affects accuracy:

Inverse Distance Weighting (IDW)

IDW estimates unknown elevations based on nearby known points, weighting closer points more heavily. Simple and fast, but can produce “bull’s-eye” patterns around data points. Best for evenly distributed, dense point clouds.

Triangulated Irregular Network (TIN)

TIN connects original data points into triangular facets, preserving the exact measured values at survey points. Excellent for engineering applications where survey accuracy must be maintained. Can be converted to raster DTM when grid format is required.

Kriging

A geostatistical interpolation method that models spatial autocorrelation. Kriging produces smooth surfaces and provides error estimates at each location. Computationally intensive but delivers high-quality results for sparse data. Preferred in scientific and research applications.

Natural Neighbor

Uses Voronoi diagrams to determine point influence areas. Produces smoother results than IDW without the bull’s-eye artifacts. Good general-purpose choice for most DTM generation tasks.

Lidarvisor uses optimized interpolation algorithms automatically selected based on point density and terrain characteristics, ensuring optimal DTM quality without manual parameter tuning.

Common DTM Quality Issues

Understanding potential DTM problems helps you evaluate data quality and troubleshoot issues:

Vegetation Artifacts

Low vegetation points misclassified as ground create elevated “bumps” in the DTM. Common under dense shrubs or crops where ground returns are scarce. Solution: Improve classification parameters or use manual editing for critical areas.

Bridge and Culvert Issues

LiDAR pulses hit bridge decks, not the terrain beneath. DTMs may show bridges as terrain ridges, disrupting drainage analysis. Some workflows use “bridge deck” classification to remove these features.

Water Surface Problems

Water absorbs infrared LiDAR, producing sparse, noisy returns over lakes and rivers. DTMs may show irregular or missing data over water bodies. Hydro-enforcement processing can flatten water surfaces to appropriate elevations.

Edge Effects

DTM quality degrades near survey boundaries where interpolation has limited neighboring points. Allow buffer zones around areas of interest when planning data collection.

DTM Use Case Examples

Real-world applications demonstrate the value of high-quality DTMs across industries:

Flood Risk Assessment

A county planning department needs to update flood hazard maps. Using drone LiDAR data processed through Lidarvisor, they generate a 0.5m resolution DTM of river corridors. The bare-earth surface feeds directly into HEC-RAS flood modeling software, producing accurate 100-year flood extent predictions for insurance rate maps and development regulations.

Highway Corridor Design

An engineering firm designing a new highway route requires DTM data for earthworks calculations. Airborne LiDAR captures the 50-mile corridor, and Lidarvisor generates a DTM with 2cm vertical accuracy. Civil engineers import the surface into Civil 3D for cut/fill optimization, saving hundreds of thousands of dollars by identifying the most efficient alignment.

Precision Agriculture Drainage

A farm operation struggles with wet spots reducing crop yields. Drone LiDAR and Lidarvisor processing produces a high-resolution DTM revealing subtle topographic depressions invisible to the eye. The farmer uses this data to plan tile drain placement, improving field drainage and increasing yields by 15% in previously waterlogged areas.

Archaeological Discovery

Researchers investigating a forested region suspected of containing ancient settlements acquire airborne LiDAR. After Lidarvisor processing, the bare-earth DTM reveals geometric patterns and linear features hidden under centuries of vegetation growth—evidence of previously unknown structures that traditional surveys never detected.

Software Tools for DTM Creation

Several software options exist for generating DTMs from point cloud data:

Cloud-Based Solutions

• Lidarvisor: Fully automated DTM generation with AI-powered classification. Upload LiDAR data and receive survey-ready DTM, hillshade, contours, and slope maps. No software installation or GIS expertise required.

Desktop GIS Software

• QGIS: Free, open-source option with LAStools and PDAL plugins for point cloud processing
• ArcGIS Pro: Comprehensive toolset for LiDAR classification and DTM generation (commercial license)
• Global Mapper: User-friendly interface for LiDAR data processing and terrain analysis

Specialized LiDAR Software

• Terrasolid: Industry standard for large-scale LiDAR processing (enterprise pricing)
• LiDAR360: Point cloud processing with classification tools
• CloudCompare: Free, open-source point cloud editing and processing

For most users, Lidarvisor offers the fastest path from raw LiDAR data to usable DTM products—without the learning curve of traditional desktop software.