ZoningKit Analysis and Overseas Deployment

What is inside the ZoningKit used by MORIZON NEXT (a QGIS plugin), and what is needed to use it outside Japan ｜ 2026-10-19

1. Purpose

To examine the contents of the ZoningKit from the Forestry Agency's "MORIZON" that MORIZON NEXT downloads, and to sort out what is needed to make it usable for forestry, forest management and environmental measures outside Japan, such as in Brazil, Canada and Spain.

2. Contents of ZoningKit

Distributed by the G-Spatial Information Center (CKAN dataset rinya-morizon-dateset). There is one ZoningKit_XX.zip for each of zones I–XIII of the Japan Plane Rectangular Coordinate System (zones XIV–XIX are missing). The zone VIII ZIP (about 423 MB) contains:

The only substantive data are the three rasters NPP, SRAD and VTEX used to calculate the site index.

Properties of the three rasters (zone VIII)

According to the manual, NPP was made from 1 km gridded climate data, and SRAD and VTEX from a 50 m elevation grid, as uniform nationwide datasets.

3. How the plugin uses them

The three rasters feed into the estimation formula for the site index (height of dominant trees at age 40, in m) (processes/raster_writer/siteidx.py).

Site index = a + (NPP − b)·c − (SRAD − d)·0.01·e − (VTEX − f)·0.01·g

Default coefficients exist for each species (settings_manager.py). Source: Mitsuda, Y. and Kitahara, F. (2017).

4. What is needed to use it in other countries

The ZoningKit part

1. Recreate NPP, SRAD and VTEX with the same definitions. SRAD and VTEX can likely be made from any country's DEM once the calculation method is known. NPP requires climate data, and the formula used must be confirmed in Mitsuda and Kitahara (2017) or Forestry Agency documents (the method cannot be determined from these files).
2. Recreate the coefficients. The coefficients were fitted to Japanese forest survey data and three Japanese species, so they cannot be used as-is for other species or climates. Either build a local site index model, or directly supply a site index raster made by another method (the manual says a "custom site index" can be specified in the input field).
3. Replace the yarding efficiency settings. They are values for Japanese harvesting systems and must be replaced with local systems.

Other Japan-specific parts besides ZoningKit

The heaviest part of overseas support is how to derive the site index, i.e., how to replace the three rasters and the Japan-specific coefficients. The rest is swapping data sources and CRS.

5. Parts that require vegetation surveys

NPP, SRAD and VTEX are made from climate and terrain, so they can be made without vegetation surveys. Specialized vegetation surveys are needed for the coefficients and species. The Japanese coefficients were obtained by fitting forest survey measurements (tree height and stand age) to the three data layers for each species.

What a survey summary needs to provide:

1. Target species: species targeted for planting and management in the region (equivalents of Japan's sugi, hinoki and karamatsu)
2. Growth reference: reference age for each species and the dominant height at that age. "Age 40" is a reference fitted to Japanese silviculture; the reference age changes for fast-growing tropical species or slow-growing boreal species
3. Georeferenced survey plots: data with tree height, stand age, species and coordinates (needed to refit the coefficients)
4. Existing materials: site class maps and yield tables. If available, they can be used by supplying a site index raster directly

6. Building it from open data

If it can be built from open data, the tool could be offered for forestry, management and environmental measures outside Japan (Brazil, Canada, Spain, etc.), expanding the countries of use. Since site index is "dominant height at a reference age", a realistic approach is to build it in three layers.

7. Status of national forest inventories in three countries

As researched on 19 October 2026.

8. Points of caution

• Different meanings of height: heights in global canopy height maps are canopy heights seen from satellites and differ in definition from the "dominant tree height" of site index. GLAD lists overestimation on mountain slopes and mixing with building heights in urban areas as known issues.
• Plantations: for fast-growing plantations such as eucalyptus in Brazil, the accuracy of global forest age maps is likely to drop. It may be necessary to supplement with domestic data such as planting-year records.
• Reference age: the Japanese reference of "age 40" needs to be redefined for each country and species.
• Reading site quality: as the manual states, a site index from a uniform nationwide formula should be read not as an absolute value but as relative quality within a region. The same applies when building it for a new region until correction accuracy is established.

Sources

• Forestry Agency, "Forest Zoning Support Tool MORIZON" ZoningKit (G-Spatial Information Center) / Site index estimation formula: Mitsuda, Y. and Kitahara, F. (2017)
• MITECO IFN4 data processing manual
• FAO Microdata: Spanish NFI4
• forestables (ecosyste.ms)
• Canada's National Forest Inventory Ground Plot Data
• Forest Inventory Ground Plots - Public Access
• Forest Inventory Ground Plot Data and Interactive Map (BC)
• SCANFI (open.canada.ca)
• SCANFI v1.2 (Earth Engine)
• IFN Brasil – Solos (dados.florestal.gov.br)
• IFN Brasil litter metadata (corrected version)
• Resolução SFB nº 23/2024
• Article on the IFN publication platform (Portal Amazônia)
• UMD GLAD GEDI global canopy height
• Mapping global forest canopy height through integration of GEDI and Landsat data
• Besnard et al. 2021, ESSD
• GAMI (B2FIND)
• ESSD 2026 vol.18 (GAMI)