SIGRID

Systematic Iterative GRID - Global Systematic Grid of points at 1x1 km with sub-nested grids at any spacing

What is it?

SIGRID was designed in order to be a Global grid where the distance between plots is homogeneous for any latitude. The plots in SIGRID are designed row by row ( East to West), moving North to South, in such a way that the latitude for each row of plots is constant and the distance in degrees that equates to 1.000 meters at that latitude is calculated and applied sequentially to each plot. This guarantees that the distance is very close to 1000 meters at any latitude.

The structure of the CSV files with SIGRID plots allows to easily choose different plot distances based on the original 1.000 meters grid. I.e. one could generate an 8km, 4km or 2km grid, maintaining ID consistency, allowing to perform nested spatial surveys for instance.

This image represents the SIGRID plots when stratified to 200km distance ( you can download the KMZ file here )

This is the detail of the SIGRID plots at 1000 m distance in an area in northern Europe where other grid types generate distortions on the spacing due to the high latitude:

How to download SIGRID

To explore the grid, as it is very large, we have created a KML file that allows for download of the grid in tiles of 10 by 10 degrees.

You can access it through this Google Earth file (KMZ)

Locate the area for which you need the grid points and then click on the tile. There will be a popup with a link to download the ZIP file containing the plots for that tile.

You can also find the list with all of grids by tile (10 by 10 degrees) directly from the Open Foris page

Each ZIP file with the points for a tile is composed of several CSV files as the number of points in the tiles close to the equator is above 1.2 million.

Data format

The CSVs that can be downloaded from the links provided in the KMZ file (that can be opened through Google Earth Pro ) have a format like this:

CE_ID	yCoordinate	xCoordinate	grid_1_SIGRID_x_20_30_y_50_60	grid_2_SIGRID_x_20_30_y_50_60	grid_3_SIGRID_x_20_30_y_50_60	grid_4_SIGRID_x_20_30_y_50_60	grid_5_SIGRID_x_20_30_y_50_60	grid_6_SIGRID_x_20_30_y_50_60	grid_8_SIGRID_x_20_30_y_50_60	grid_9_SIGRID_x_20_30_y_50_60	grid_10_SIGRID_x_20_30_y_50_60	grid_12_SIGRID_x_20_30_y_50_60	grid_15_SIGRID_x_20_30_y_50_60	grid_16_SIGRID_x_20_30_y_50_60	grid_20_SIGRID_x_20_30_y_50_60	grid_25_SIGRID_x_20_30_y_50_60	grid_30_SIGRID_x_20_30_y_50_60	grid_50_SIGRID_x_20_30_y_50_60	grid_100_SIGRID_x_20_30_y_50_60
2790_8986	59.9949504	29.9851232	true	true	false	false	false	false	false	false	false	false	false	false	false	false	false	false	false
2790_8987	59.9949504	29.9672064	true	false	false	false	false	false	false	false	false	false	false	false	false	false	false	false	false
2790_8988	59.9949504	29.9492864	true	true	true	false	false	true	false	false	false	false	false	false	false	false	false	false	false
2790_8989	59.9949504	29.9313664	true	false	false	false	false	false	false	false	false	false	false	false	false	false	false	false	false
2790_8990	59.9949504	29.9134496	true	true	false	false	true	false	false	false	true	false	false	false	false	false	false	false	false
2790_8991	59.9949504	29.8955296	true	false	true	false	false	false	false	true	false	false	false	false	false	false	false	false	false

The column CE_ID is the unique ID of the plot, composed of two number Y_X. The Y represents the row from the starting point of the grid generation 85 degrees North, 169 degrees West (i.e. 2000_3300 means 2.000 km south of the starting latitude of the grid - which is 85 degrees North - and 3.300 km West of the starting longitude - which is 169 degrees West ).

This is followed by the Latitude (yCoordinate) and Longitude (xCoordinate) of the center of the plot.

The columns that follow indicate to which subgrid the plot belongs to . All the plots belong to the subgrid 1 (1000 m distance) while 1/4 of the plots belong to the 2x2 km subgrid, 1/9 of th plots belong to the 3x3km subgrid, 1/16 of the plots belong to the 4x4 km subgrid, 1/25 of the plots belong to the 5x5 km grid and so on. The information of the subgrid columns can be combined to filter the plots into any desired subgrid. I.e. the subgrid of plots for 18 x18 km can be obtained combining plots that belong to both the 9x9 km subgrid and the 6x6km subgrid ( using the filters in tools like Excel).

Having a homogeneus grid like SIGRID allows to generate Collect Earth assessments that can be easily intensified. You could start with an 8x8 km grid and depending on your resources or accuracy goals on the assessment easily move on to the remaining plots in the 4x4 km grid that were not assessed already. Again, you could move to the 2x2km grid following the same logic and even to a 1x1 km grid to achieve maximum accuracy.

Download SIGRID by country

The grid is also published pre-cut by country, so you don't need to download the 10x10 degree tiles and clip them yourself. There is one ZIP file per country, available from the Open Foris page.

The easiest way to browse them is through Google Earth: open the SIGRID country grids file (KMZ) (also found in the resources folder of this project), click on the placemark of a country and use the download link in the popup.

Each country ZIP contains:

• <Country>_1x1km.csv — the full 1x1 km grid for the country, with the same subgrid columns as the tile format described above. In addition, every plot is tagged with the province (ADM1NM) and district (ADM2NM) it falls in, based on the UN 2023 administrative boundaries.
• <Country>_2x2km.csv, <Country>_3x3km.csv ... <Country>_100x100km.csv — one CSV per sampling density. These are exact subsets of the 1x1 km file (the CE_IDs are consistent), so an assessment started at a coarse density can later be intensified with the finer files, as described above.
• <Country>_plot_counts_by_district.csv — the number of plots per province/district at each density.
• A README.txt describing how the files were generated.

For some countries the UN 2023 boundaries do not include district names; those are published as <Country>_grids_no_districts.zip and their plots carry no province/district columns.

Note for Collect Earth users: Google Earth degrades with files above ~2,000 placemarks, so for larger files use Tools → Utilities → Divide large CSV files in Collect Earth (aggregating by the ADM1NM or ADM2NM column works well).

How to reproduce the grid

You need a Java JDK (8 or newer) and Maven. Clone the repository, build it and run the main class of GenerateSigrid.java (code):

git clone https://github.com/herrtunante/SIGRID.git
cd SIGRID
mvn clean compile
mvn exec:java -Dexec.mainClass=org.openforis.sigrid.GenerateSigrid

You can also simply run the main method of GenerateSigrid from your IDE. Either way, run it with the repository root as the working directory: the code reads the SQL scripts in resources/ through relative paths and writes its results to an output/ folder that it creates there.

By default this generates the plots for the whole globe (85°N to 85°S) and writes them as one zipped CSV per sub-grid density into output/ (global_1000m_1_subgrid.csv.zip, global_1000m_2_subgrid.csv.zip ... global_1000m_100_subgrid.csv.zip). Files with more than 500,000 plots are split into several CSVs inside the ZIP.

Generating only a region

Generating all of the plots worldwide can take up to one day on a standard computer, so there is an option to limit the plots to those within a bounding box. In the main method of GenerateSigrid, comment out globalGrid.generate() and use the call that takes a bounding box, specifying the East, North, West and South limits in decimal degrees:

// East, North, West, South
globalGrid.generate( 40d, 20d, 30d, 10d ); // plots between longitudes 30-40 and latitudes 10-20

A 10x10 degree box like this one (about 1.2 million plots) generates in under a minute, because entire rows outside the latitude range are skipped.

Storing the plots in a database

The STORE constant in GenerateSigrid selects where the plots go. The default is CSVStore (the CSV files described above); change it to new JDBCStore() to write into a database instead. JDBCStore uses an SQLite file (sigrid.db in the working directory) by default; to use PostgreSQL instead, set USE_SQLITE to false in JDBCStore.java and fill in the connection URL and credentials there.

Once the plots are in a database, QuerySigrid.java (code) can export them without regenerating anything:

• writeCsvFromBoundingBox(...) exports the plots of any bounding box and sub-grid to CSV — the main method contains many commented examples for individual countries that you can adapt.
• generateTiledGridsAll() exports the 10x10 degree tiles published on the Open Foris server.
• generateKmlGrids(...) regenerates the tile-browser KML from the FreeMarker template in resources/kml_template.fmt.

mvn exec:java -Dexec.mainClass=org.openforis.sigrid.QuerySigrid

Note: the layout of the gridflags bitmask column changed in 2026 (the previous encoding could not represent the 50 km and 100 km sub-grids). A database populated with the old code can be converted in place by running resources/migrateGridflags.sql against it.