In the remote sensing industry, I frequently encounter confusion regarding some of the most basic terminology for 3-D digital models—in particular, the digital elevation model (DEM) versus the digital terrain model (DTM). Many people use these terms interchangeably, but there are key differences that have to do with both file type and content.

There are three types of models that are most common to our industry: the DEM, the DTM, and the digital surface model (DSM). The DSM is rarely confused with the others as this model captures the reflective surface heights, such as buildings, power lines, and tree canopy. The DEM and the DTM, however, are both “bare earth” models, capturing only ground elevation data.

A DTM contains vector-based source information such as discrete LiDAR measurements or photogrammetric mass points and breaklines. A DEM is a raster-based product with a single unique elevation measurement representing the area covered by each pixel of the file. A DEM is most often created directly from a DTM source by applying algorithmic processes and it reflects some level of generalization and normalization when compared to the DTM. Another way of looking at it is that a DEM has gridded elevation points at regularly-spaced intervals in x and y, whereas a DTM has irregularly-spaced elevation points and breaklines.

Often, when someone requests that a new DEM be developed for their project, they are actually seeking a DTM. Considering both the file type and the level of detail, there are critical differences between a DEM and a DTM. When ordering or supplying, it is important to state to your service provider or client exactly what you are expecting to receive or supply. Stating the wrong data type could be a costly mistake for all involved parties.