Skip to contents

Temporal-spatial Data

Data structures

Macro dimensions (Dim)

From a macro perspective, TimeSpatData is always defined in three dimensions:

  • Time: timestamps
  • Spat: spatial
  • Vari: multiple variables

However, the Spat dimension actually has a different and complex form, as it is based on geological data raster and vector.

In the vertical direction, we only support the layer dimension in this version. This is not the physical vertical dimension like x and y, but rather an extracted layer as a homogeneous layer.

Data (array) dimensions (dim)

During the data-processing and -storage, the data is always given in array form, the macro dimensions cannot be clearly defined in an array dimension, therefore the array dimensions are redefined:

time

Dimension time is defined for time and has backward consistency. It always stands in the first dimension, with two possible options:

  • continuous: With a fixed time step, defined by units and original_time
  • discrete: Without the concept of time step, but always giving the exact timestamps

spat in horizontal

Dimension spat is defined for vector spatial data (points, lines, and polygons). It is not ordinal, but is linked to the Spat_ID and Spat_Data variables. The coordinate reference system (CRS) is defined using an EPSG-code.

x und y in horizontal

Die beide Dimensionen sind für Raster (einzeln Schichten) horizontal Koordinate definiert mit die Koordinaten-Werten von zentralen Punkten des Zells. Das CRS ist mit EPSG-code definiert. Deshalb ist es nicht erlaubt die Einheit oder andere zusätzliche Sachen von den beiden Dimensionen zu definieren.

The two dimensions x and y are defined for raster data in single layers. They represent the horizontal coordinates of the central points of cells in the raster. The coordinate reference system (CRS) is defined using an EPSG-code. It is not allowed to define units or other additional information for these dimensions.

layer in vertical

Dimension layer is defined as a conceptual layer without an exact vertical position.

vari

Dimension vari is defined for multiple corresponding variables with Name and Unit. The dimension always stands last.

Data type (name)

Spat

Due to the variety of spatial data forms, there are four Spat-Form listed below:

  • Vect [spat]: Vector data (points, lines, polygons)
  • VectLayer [spat, layer]: Vector data with multiple layers
  • Rast [x, y]: Raster data in a single layer
  • RastLayer [x, y, layer]: Raster data with multiple layers

Vari

Due to the different properties of variables, the following four Vari-Form are defined:

  • Vari [vari = 1] / []: single variable
  • Array [vari]: multiple variables with the same data-structure and -size

In summary, a complete TimeSpatData must define the three macro dimensions (not data dimensions).

TimeSpatData in NetCDF

The definition of dimensions in NetCDF remains the same as the data structure dimensions described above. Due to the limits of NetCDF, it is also necessary to further define other additional dimensions, variables, and attributes.

Processing Tools

Under the TSD structure, there are four main tools: read, write, crop, and extract. For more details, see the section read_tsd(), write_tsd(), crop_tsd() and extract_tsd().