17. Working with Mesh Data

17.1. What’s a mesh?

A mesh is an unstructured grid usually with temporal and other components. The spatial component contains a collection of vertices, edges and faces in 2D or 3D space:

  • vertices - XY(Z) points (in the layer’s coordinate reference system)

  • edges - connect pairs of vertices

  • faces - a face is a set of edges forming a closed shape - typically a triangle or a quadrilateral (quad), rarely polygons with more vertices


Fig. 17.1 Different mesh types

QGIS can currently render mesh data using triangles or regular quads.

Mesh provides information about the spatial structure. In addition, the mesh can have datasets (groups) that assign a value to every vertex. For example, having a triangular mesh with numbered vertices as shown in the image below:


Fig. 17.2 Triangular grid with numbered vertices

Each vertex can store different datasets (typically multiple quantities), and those datasets can also have a temporal dimension. Thus, a single file may contain multiple datasets.

The following table gives an idea about the information that can be stored in mesh datasets. Table columns represent indices of mesh vertices, each row represents one dataset. Datasets can have different datatypes. In this case, it stores wind velocity at 10m at a particular moments in time (t1, t2, t3).

In a similar way, the mesh dataset can also store vector values for each vertex. For example, wind direction vector at the given time stamps:

10 metre wind




10 metre speed at time=t1




10 metre speed at time=t2




10 metre speed at time=t3




10m wind direction time=t1




10m wind direction time=t2




10m wind direction time=t3




We can visualize the data by assigning colors to values (similarly to how it is done with Singleband pseudocolor raster rendering) and interpolating data between vertices according to the mesh topology. It is common that some quantities are 2D vectors rather than being simple scalar values (e.g. wind direction). For such quantities it is desirable to display arrows indicating the directions.


Fig. 17.3 Possible visualisation of mesh data

17.2. Supported formats

QGIS accesses mesh data using the MDAL drivers. Hence, the natively supported formats are:

  • NetCDF: Generic format for scientific data

  • GRIB: Format commonly used in meteorology

  • XMDF: As an example, hydraulic outputs from TUFLOW modelling package

  • DAT: Outputs of various hydrodynamic modelling packages (e.g. BASEMENT, HYDRO_AS-2D, TUFLOW)

  • 3Di: 3Di modelling package format based on Climate and Forecast Conventions (http://cfconventions.org/)

  • Some examples of mesh datasets can be found at https://apps.ecmwf.int/datasets/data/interim-full-daily/levtype=sfc/

To load a mesh dataset into QGIS, use the addMeshLayer Mesh tab in the Data Source Manager dialog. Read Loading a mesh layer for more details.

17.3. Mesh Dataset Properties

17.3.1. Information Properties


Fig. 17.4 Mesh Layer Properties

The metadata Information tab is read-only and represents an interesting place to quickly grab summarized information and metadata on the current layer. Provided information are:

  • general such as name in the project, source path, list of auxiliary files, last save time and size, the used provider

  • based on the provider of the layer: extent, vertex, face, edges and/or dataset groups count

  • the Coordinate Reference System: name, units, method, accuracy, reference (i.e. whether it’s static or dynamic)

  • extracted from filled metadata: access, extents, links, contacts, history…

17.3.2. Source Properties

The Source tab displays basic information about the selected mesh, including:


Fig. 17.5 Mesh Layer Source Properties

  • the layer name to display in the Layers panel

  • setting the Coordinate Reference System: Displays the layer’s Assigned Coordinate Reference System (CRS). You can change the layer’s CRS by selecting a recently used one in the drop-down list or clicking on setProjection Select CRS button (see Coordinate Reference System Selector). Use this process only if the CRS applied to the layer is wrong or if none was applied.

  • The Available datasets frame lists all the dataset groups (and subgroups) in the mesh layer, with their type and description in a tree view. Both regular datasets (i.e. their data is stored in the file) and virtual datasets (which are calculated on the fly) are listed.

    • Use the add Assign extra dataset to mesh button to add more groups to the current mesh layer.

    • collapseTree Collapse all and expandTree Expand all the dataset tree, in case of embedded groups

    • If you are interested in few datasets, you can uncheck the others and make them unavailable in the project

    • Double-click over a name and you can rename the dataset.

    • refresh Reset to defaults: checks all the groups and renames them back to their original name in the provider.

    • Right-click over a virtual dataset group and you can:

      • Remove dataset group from the project

      • Save dataset group as… a file on disk, to any supported format. The new file is kept assigned to the current mesh layer in the project.

  • Checking the unchecked Treat as static dataset group allows to ignore the map temporal navigation properties while rendering the mesh layer. For each active dataset group (as selected in symbology Symbology ► general Datasets tab), you can:

    • set to None: the dataset group is not displayed at all

    • Display dataset: eg, for the “bed elevation” dataset which is not time aware

    • extract a particular date time: the dataset matching the provided time is rendered and stay fixed during map navigation.

17.3.3. Symbology Properties

Click the symbology Symbology button to activate the dialog. Symbology properties are divided into several tabs: Datasets

The tab general Datasets is the main place to control and set which datasets will be used for the layer. It presents the following items:

  • Groups available in the mesh dataset, with whether they provide:

    • meshcontoursoff scalar dataset

    • or meshvectorsoff vector dataset: by default, each vector dataset has a scalar dataset representing its magnitude automatically generated.

    Click on the icon next to the dataset name to select the group and type of data to represent.

  • Selected dataset group(s) metadata, with details on:

    • the mesh type: edges or faces

    • the data type: vertices, edges, faces or volume

    • whether it’s of vector type or not

    • the original name in the mesh layer

    • the unit, if applicable

  • blending mode available for the selected datasets.


Fig. 17.6 Mesh Layer Datasets

You can apply symbology to the selected vector and/or scalar group using the next tabs. Contours Symbology


The meshcontours Contours tab can be activated only if a scalar dataset has been selected in the general Datasets tab.

In the meshcontours Contours tab you can see and change the current visualization options of contours for the selected group, as shown in Fig. 17.7 below:


Fig. 17.7 Styling Contours in a Mesh Layer

  • Use the slider or the spinbox to set the Opacity of the current group.

  • Use refresh Load to adjust the min and max values of the current group or enter custom values if you want to exclude some.

  • Select the Resampling method

  • Classify the dataset using the color ramp shader classification. Vectors Symbology


The meshvectors Vectors tab can be activated only if a vector dataset has been selected in the general Datasets tab.

In the meshcontours Vectors tab you can see and change the current visualization options of vectors for the selected group, as shown in Fig. 17.8:


Fig. 17.8 Styling Vectors in a Mesh Layer with arrows

Mesh vector dataset can be styled using various types of Symbology:

  • Arrows: vectors are represented with arrows at the same place as they are defined in the raw dataset (i.e. on the nodes or centre of elements) or on a user-defined grid (hence, they are evenly distributed). The arrow length is proportional to the magnitude of the arrow as defined in the raw data but can be scaled by various methods.

  • Streamlines: vectors are represented with streamlines seeded from start points. The seeding points can start from the vertices of the mesh, from a user grid or randomly.

  • Traces: a nicer animation of the streamlines, the kind of effect you get when you randomly throws sand in the water and see where the sand items flows.

Available properties depend on the selected symbology as shown in the following table.


Description and Properties




Line width

Width of the vector representation




Coloring method

  • a Single color assigned to all vectors

  • or a variable color based on vectors magnitude, using a Color ramp shader




Filter by magnitude

Only vectors whose length for the selected dataset falls between a Min and Max range are displayed



Display on user grid

Places the vector on a grid with custom X spacing and Y spacing and interpolates their length based on neighbours



Head options

Length and Width of the arrow head, as a percentage of its shaft length


Arrow length

  • Defined by Min and Max: You specify the minimum and maximum length for the arrows, QGIS will interpolate their size based on the underlying vector’s magnitude

  • Scale to magnitude: arrow length is proportional to their vector’s magnitude

  • Fixed: all the vectors are shown with the same length


Streamlines seeding method

  • On mesh/grid: relies on the user grid to display the vectors

  • Randomly: vector placement is randomly done with respect to a certain density


Particles count

The amount of “sand” you want to throw into visualisation


Max tail length

The time until the particle fades out

checkbox Rendering

In the tab meshframe Rendering tab, QGIS offers possibilities to display the grid, as shown in Fig. 17.9:

  • Native Mesh Rendering that shows original faces and edges from the layer

  • Triangular Mesh Rendering that adds more edges and displays the faces as triangles


Fig. 17.9 Mesh Rendering

The lines width and color can be changed in this dialog, and both the grid renderings can be turned off. Stacked mesh averaging method

3D layered meshes consist of multiple stacked 2D unstructured meshes each extruded in the vertical direction (levels) by means of a vertical coordinate. The vertices and faces have the same topology in each vertical level. Values are usually stored on the volumes that are regularly stacked over base 2d mesh. In order to visualise them on 2D canvas, you need to convert values on volumes (3d) to values on faces (2d) that can be shown in mesh layer. The meshaveraging Stacked mesh averaging method provides different averaging/interpolation methods to handle this.

You can select the method to derive the 2D datasets and corresponding parameters (level index, depth or height values). For each method, an example of application is shown in the dialog but you can read more on the methods at https://fvwiki.tuflow.com/index.php?title=Depth_Averaging_Results.

17.3.4. Temporal Properties

The temporal Temporal tab provides options to control the rendering of the layer over time. It allows to dynamically display temporal values of the enabled dataset groups. Such a dynamic rendering requires the temporal navigation to be enabled over the map canvas.


Fig. 17.10 Mesh Temporal properties

Layer temporal settings

  • Reference time of the dataset group, as an absolute date time. By default, QGIS parses the source layer and returns the first valid reference time in the layer’s dataset group. If unavailable, the value will be set by the project time range or fell back to the current date. The Start time and End time to consider are then calculated based on the internal timestamp step of the dataset.

    It is possible to set a custom Reference time (and then the time range), and revert the changes using the refresh Reload from provider button.

  • Dataset matching method: determines the dataset to display at the given time. Options are Find closest dataset before requested time or Find closest dataset from requested time (after or before).

Provider time settings

  • Time unit extracted from the raw data, or user defined. This can be used to align the speed of the mesh layer with other layers in the project during map time navigation. Supported units are Seconds, Minutes, Hours and Days.

17.3.5. Metadata Properties

The editMetadata Metadata tab provides you with options to create and edit a metadata report on your layer. See Metadata for more information.

17.4. Mesh Calculator

The Mesh Calculator tool from the top Mesh menu allows you to perform arithmetic and logical calculations on existing dataset groups to generate a new dataset group (see Fig. 17.11).


Fig. 17.11 Mesh Calculator

The Datasets list contains all dataset groups in the active mesh layer. To use a dataset group in an expression, double click its name in the list and it will be added to the Mesh calculator expression field. You can then use the operators to construct calculation expressions, or you can just type them into the box.

The Result Layer helps you configure properties of the output layer:

  • checkbox Create on-the-fly dataset group instead of writing layer to disk:

    • If unchecked, the output is stored on disk as a new plain file. An Output File path and an Output Format are required.

    • If checked, a new dataset group will be added to the mesh layer. Values of the dataset group are not stored in memory but each dataset is calculated when needed with the formula entered in the mesh calculator. That virtual dataset group is saved with the project, and if needed, it can be removed or made persistent in file from the layer Source properties tab.

    In either case, you should provide a Group Name for the output dataset group.

  • The Spatial extent to consider for calculation can be:

    • a Custom extent, manually filled with the X min, X max, Y min and Y max coordinate, or extracted from an existing dataset group (select it in the list and press Use selected layer extent to fill the abovementioned coordinate fields)

    • defined by a polygon layer (Mask layer) of the project: the polygon features geometry are used to clip the mesh layer datasets

  • The Temporal extent to take into account for datasets can be set with the Start time and End time options, selected from the existing dataset groups timesteps. They can also be filled using the Use all selected dataset times button to take the whole range.

The Operators section contains all available operators. To add an operator to the raster calculator expression box, click the appropriate button. Mathematical calculations (+, -, *, … ) and statistical functions (min, max, sum (aggr), average (aggr), … ) are available. Conditional expressions (=, !=, <, >=, IF, AND, NOT, … ) return either 0 for false and 1 for true, and therefore can be used with other operators and functions. The NODATA value can also be used in the expressions.

The Mesh Calculator Expression widget shows and lets you edit the expression to execute.