A tiled grid dataset is a decomposition of a grid dataset into separate files describing a square portion of space. These files are indexed in a file structure allowing to retrieve them by geographical location.
Examples of tiled grids are available in this repository. See for example this one.
A tiled grid dataset is composed of the following elements:
The tiled grid is described by a info.json file at the root of the dataset structure. See an example of such info.json file here or below:
{
"dims": [
"tcd",
"dlt"
],
"crs": "3035",
"tileSizeCell": 128,
"originPoint": {
"x": 0,
"y": 0
},
"resolutionGeo": 500,
"tilingBounds": {
"yMin": 14,
"yMax": 84,
"xMax": 114,
"xMin": 14
}
}
This info.json file includes the following information:
| Entry | Type | Description |
|---|---|---|
| dims | Array(String) | The list of dimension column names, corresponding to the grid variables. |
| crs | String | The EPSG code of the grid Coordinate Reference System. |
| tileSizeCell | Integer | The size of each tile, in number of cells. The tiles are square. They all have the same size. There is no constraint on this value: It is not forced to be a power of 2. It should be chosen to ensure the tile size is low enough (so that it is fast enough to download and parse) and high enought (so that there are not too many tiles to be requested). |
| originPoint | { x:Number, y:Number } | The geographical coordinates (in grid CRS) of the lower left point of the grid. This is the bottom left point. The tiles start being built and numbered from this point. |
| resolutionGeo | Number | The resolution of the grid, that is the size of each cell expressed in the ground CRS unit of measure (usually meters). |
| tilingBounds | { xMin:Number, yMin:Number, xMax:Number, yMax:Number } | The bounds of the tiling scheme, that is the min/max x/y coordinates of the tiles in the tiling scheme. tilingBounds refers to how the tiling frame is filled with data: Not all tiles may be provided, due to data sparsity. tilingBounds may be used to avoid requesting tiles where no tile is expected to be found. |
The information provided in this info.json file allows to retrieve each tile by geographic location and interpret the information of each tile file to locate properly the cells it contains.
The tiles are structured into a folder structure following the http://…/X/Y.fff pattern. X and Y are the coordinates of the tiles within the tiling scheme, starting with (0,0) for the tile immediately on the top right of the origin point. fff is the file extension of each file, depending on the chosen encodings.
See for example this dataset.
Each tile file contains the information related to the grid for the geographical location of the tile. It is a simple tabular file. Each row corresponds to a grid cell. The following columns are provided:
Each file can be encoded in whatever user defined format. GridViz currently supports CSV and Apache Parquet. Other encodings could be considered.
This format is designed for the specific need of gridded data visualisation on the web. It uses a tiling mechanisms like other similar formats.
Difference with slippy map /Z/X/Y.png tilenames:
Difference with Cloud Optimized GeoTIFF (COG)