CARTEMA
Open Cartema

For editors and owners

Importing data

The formats Cartema reads, uploading a CAD drawing (DXF, DWG), where the new layers go, coordinate systems, rasters and terrain models, field settings, and replacing data with a delivery.

Updated 2026-11-03For editors and owners of a workspace

Everything on a map starts as a dataset: a table of features, or an image. Studio → Datasets holds them all, in folders and views — see Your data.

Importing a file

In Studio → Datasets, Import a file opens the import: give the dataset a Name, choose the File, choose where its layers go (below), and Import. The import runs in the background — its progress and log are in Jobs — and the dataset is Ready when it has finished.

Where the new layers go

Under Add to apps, tick each app the new layers should appear in and choose a folder of its layer tree: the top of the tree, one of its folders (folders inside folders are listed by their path, for example Plans / 2026), or New folder… — give it a name and choose the folder it goes Inside. If that folder already has a folder of the same name, the layers go into it rather than into a second one.

When the import has finished, each layer is added at the top of the chosen folder, drawn with the dataset's default style, as a new version of the app: it shows in the app's History under your name, and Restore brings back the version before it (see Building apps in Studio). A file that becomes several datasets (a drawing's lines, areas and text, a GeoPackage's layers) adds all of them together.

With no folder chosen (the top of the tree), each layer goes where the app's layer tree keeps its kind: a CAD drawing into Plans, in a folder of its own named after the file (Plans is made just before the basemaps folder when the app has none); an image or terrain model on top of the basemaps folder; any other layer into Custom layers (made first when missing). A folder you choose is always kept. In an app that already shows something when it opens, the new layers start switched off: switch them on in the map's legend, or in the app's layer tree in Studio. The same holds for a file uploaded from an app's map into the app, and for a personal layer added to an app.

  • Only the apps you may change are listed. Your access is checked again when the layers are added: an app you can no longer change is left as it was.
  • An app open only to listed people keeps to them — the new layers are seen by exactly the people who open the app.
  • If a layer cannot be added to an app (for example because its folder was deleted in the meantime), the import is still complete: the dataset is in Studio → Datasets, and Jobs says which app it could not be added to. Add it from the app's layers, or with Add to app….
  • Tick no app to import the data only.

Formats

Kind Formats
Vector GeoJSON, GeoPackage, Shapefile (zipped), File Geodatabase (zipped), KML / KMZ, GML, FlatGeobuf, SpatiaLite, CSV, DXF and DWG (CAD drawings, AutoCAD R13 to 2018)
Imagery GeoTIFF, JPEG 2000, ERDAS Imagine, MrSID (.sid) on request; PNG or JPG only zipped together with their world file and .prj
Terrain GeoTIFF, SRTM .hgt, USGS .dem, .asc, .xyz, and .bil / .flt with their .hdr (zipped)
  • One dataset per zip. A zip with several shapefiles is refused — upload them one at a time.
  • A file with several layers (a GeoPackage, a geodatabase, a DXF) is imported as one dataset per layer, unless you name one in Layer.
  • CSV: point coordinates are found in columns named like x / lon / longitude / easting and y / lat / latitude / northing, or geometry in a column named wkt / geom / geometry.
  • Spreadsheets — Excel (.xlsx, .xlsm, .xls) and OpenDocument (.ods) — are read by My layers → Upload in an app's map and by Load into this layer: they take a workbook, and each sheet with X and Y (or longitude and latitude) columns becomes a layer of points; sheets without them are left out. Import a file here takes CSV, so save the sheet as CSV first (see Loading a file into a layer).
  • Images and terrain are turned into cloud-optimised GeoTIFF, so they draw fast at any zoom.
  • An orthophoto is rarely a perfect rectangle, and the empty area around the picture is often filled with plain white or black. Cartema makes that border see-through, so the map and the images next to it show through. Only the plain area joined to the image's edge is removed; white roofs or dark water inside the picture stay. Images already in Cartema get the same treatment automatically, a few at a time, so an old image may lose its border a little while after you first notice it.
  • ECW images cannot be read: export them as GeoTIFF or JPEG 2000.
  • A file can be up to 512 MB, and an image or terrain model (GeoTIFF, JPEG 2000 and the like) up to 8 GB — an orthophoto processed elsewhere goes in this way; for larger files, contact Cartema support.

Coordinate systems

Cartema reads the coordinate system from the file, or from the .prj that comes with it. Some files carry none — CAD drawings, CSV, images with only a world file. For those, choose the Coordinate system on the import form (for example a UTM zone or a national grid), or zip the file together with its .prj. A file whose coordinates are not in degrees and whose system is not given is refused, with a message saying so.

A CAD drawing usually carries no coordinate system, so Cartema looks for it in the grids the app uses. The upload window in an app's map chooses the app's grid as soon as a .dxf or .dwg is picked; change it if the drawing is in another. When no system is chosen at all, Cartema tries the grids the app shows, then those of the workspace's other apps:

  • If the drawing's coordinates lie in exactly one of them, it is read in that grid, and the upload says which one and why. Wrong? Choose another reads it again in the system you pick; the layers of the first reading are then taken out of the app, unless they have been edited or added elsewhere since.
  • If they fit several grids that would put it in different places, nothing is guessed: the upload lists those grids, and one click reads the drawing in the one you choose.

An image or terrain model must say where it is. One without a coordinate system (and none chosen on the form or given in a .prj), one whose system says latitude and longitude while its numbers are metres, or one that would cover the whole world, is refused: the message names the file and asks for its coordinate system.

CAD drawings (DXF, DWG)

A drawing is shown as a plan. It comes in as up to four datasets — lines, areas (hatches, filled shapes, and every closed polyline, which also stays among the lines), text and points — each styled as drawn: the CAD layers with their colours, switchable one by one in the legend; line weights and line types (dashed, centre, hidden); texts at their drawn height, which scale with the map, turned and anchored as written; blocks exploded into their parts. Each dataset's fields carry the CAD layer (cad_layer), the entity's colour and width, and for text its height in metres; the style editor can change any of it like any other style.

To upload a CAD file:

  1. In Studio → Datasets, choose Import a file.
  2. Give the dataset a Name, and choose the .dxf or .dwg as the File — or a zip of the drawing with its .prj.
  3. Without a .prj, choose the drawing's Coordinate system (for example a UTM zone or a national grid).
  4. Under Add to apps, tick the apps whose layer tree it goes into, and the folder; tick none to import the data only.
  5. Choose Import. Jobs shows its progress, and the drawing arrives as its lines, areas, text and points, drawn as in the drawing.

A DXF or DWG is read wherever a file comes in: Import a file here, My layers → Upload in an app's map, and Load into this layer, which puts a drawing's lines, areas, text or points into an existing layer. A drawing needs its coordinate system: a .prj beside it in a zip, chosen on the upload form, or the app's grid (see Coordinate systems). A DWG is converted to DXF first (AutoCAD R13 to 2018). To replace a drawing, import it again and swap the layers in the app; a delivery does not apply to drawings.

A dataset's page

In Your data, choose a dataset and Fields and settings to open its page.

Field settings for a layer of features:

Column What it sets
Label The name users read, in any language
Show Shown when a feature is identified, and in the table
Search Searched by the Search tool (text fields). With none ticked, every text field of the layer is searched; tick some to search only those. Search reads at most 64 fields, and the page names any it leaves out
Rules Required, Length, Default, Allowed values — enforced when data is edited

You can also Add field here, on any layer. On a layer made in Cartema you can remove any field; on a delivered layer only those added in Cartema, marked Added in Cartema: updates from the source keep them, with their values (see Adding a field from the table).

Value on click for an image or terrain model: whether a click on the map reads the value under it (turn it off for photographs), what the value is called (for example Altitude) and its units (m).

Use as 3D ground for a raster of heights: whether it is laid as the ground under 3D models on a tilted map (see Apps in Studio). Automatic uses it when it is named like a terrain model (DTM, DEM) and its heights hold no buildings or trees. Cartema checks this when the raster is imported: it reads a few parts of it at a couple of metres a pixel and counts the steps between neighbouring pixels. Ground slopes; a roof's edge or a tree's crown drops by several metres at once. A file called "DTM" that turns out to hold buildings is a surface model, and the page says so, with the share of steps it found. It is then left out of the ground, and the world terrain model is used there instead. Choose No to leave out any raster, or Yes to use one that the check got wrong (bare ground in steep, rocky terrain, for example). In each app, a raster is laid as the ground only where the app's layer list ticks Everyone exports for its layer, because the ground hands out its heights. The setting changes only the 3D ground. Clicks, measurements and the raster's own layer stay as they are.

Load into this layer, at the top of a layer's page, adds the features of a file to it — or replaces them — once you have matched the file's fields to the layer's; the rows that could not be loaded are listed. It reads shapefiles, KML, GeoJSON, GeoPackage, CSV and Excel (points from two columns), DXF and DWG — see Loading a file into a layer.

New layer, at the top of Studio → Datasets, makes an empty layer to draw in or to load a file into — see Layers and editing.

Replacing data: deliveries

When new versions of your data arrive — a new cadastre, an updated survey — Datasets → Deliveries → New delivery replaces the data under existing datasets while keeping their names, field labels, styles and apps. Cartema matches each file to the dataset it replaces and shows what will happen: Clean replacements, Warnings (for example a field that disappeared and is used by a style), Blocked changes, and New layers. Accept what is right and Apply. A delivery can be reverted for 14 days.

Deliveries replace the tables of delivered layers. Layers made in Cartema, and editable copies, are never touched by a delivery.