Data Setup Hints

About scipy (required for Flow Buffer / SHC computation)

Windows (standard installer): scipy is normally already included. If it's genuinely missing, close QGIS and run in Command Prompt from your QGIS install's bin folder (e.g. C:\Program Files\QGIS 3.44\bin\):
python-qgis-ltr.bat -m pip install --user scipy
(The filename may be python-qgis.bat depending on your install; adjust the path to match your QGIS version.)

Linux: close QGIS and run in a terminal:
sudo apt install python3-scipy

DEM / DSM/DTM Data

The plugin prepares a square, metre-based analysis DEM automatically, then aligns DSM to its final grid. Suitable projected CRSs are retained; geographic and Web Mercator DEMs are converted to local UTM. Keep the source CRS defined and use the same height datum for DEM and DSM. Source files are preserved; changing the project display CRS is not required. Preparation details are recorded in params.json.

• Select a general-purpose DEM, or one covering the area where you'll run the terrain analysis.
• Select a DSM covering the same extent as the DEM data.
• The DSM is aligned to the final analysis DEM grid. Valid DSM cells use canopy-height-based runoff and velocity coefficients; missing cells use the editable fallback coefficients. Read/alignment failures offer Retry, Continue without DSM, or Cancel.
• Analysis Conditions shows DSM coverage and any manual fallback. Applied settings and the actual DSM path are saved in params.json. Older runs without these records show that DSM usage is not recorded.

Virtual Shizuoka Features

• When using Virtual Shizuoka data, first pan or zoom the Design Preview to the area you want to analyze, then select VS LP/Grid DTM or VS LP -> DSM. The plugin fetches data for the preview extent at the time of selection.
• The fetched Virtual Shizuoka data extent is not continuously shown as a boundary in the preview. When the analysis area extends beyond the fetched VS LP/Grid DTM, the plugin asks whether to fetch it again (see Re-fetching before analysis below); VS LP -> DSM is fetched again together with it.
• Local DEM/DSM files can be analyzed with partial overlap, using the available data. For Virtual Shizuoka data, fetching the data for the intended analysis range first avoids re-fetching at analysis time.
• VS LP -> DSM downloads LAS files (~237 MB per tile) into vs_lp_ground/ under the analysis output folder. There is no time-based expiry: after each successful VS LP -> DSM run only the LAS tiles used by that run are kept and older LAS files are removed. Re-running the same area reuses the cached LAS without re-downloading. Delete vs_lp_ground/ manually to reclaim space.
• Fetched DEM/DSM files are cached in dem/, vs_lp_grid/ and remote_dem/. Files referenced by a saved analysis or currently loaded are kept; once no longer referenced, cache cleanup removes them when the analysis list is next refreshed.

Open data outside Japan (Netherlands / France / New Zealand)

• The DEM / DSM browser lists these sources only when the Design Preview overlaps the country: AHN4 DTM / DSM (Netherlands, 0.5 m), IGN LiDAR HD MNT / MNS (France, 0.5 m) and LINZ DEM (New Zealand, 1 m, DEM only).
• The DEM is fetched for the current Design Preview extent. For the Netherlands and France, the DSM from the same survey is then fetched automatically for the DEM extent.
• The DSM row can be cancelled, cleared and fetched again on its own. A DSM chosen later is also fetched for the DEM extent, so moving the map does not shift it. Clearing the DEM also clears the matching DSM.
• Fetched files are saved in remote_dem/ under the analysis output folder. When a file fetched earlier from the same source already covers the needed extent, it is used without downloading again. Cached DEM/DSM files are kept while an analysis references them or they are currently loaded. After they are no longer referenced, cache cleanup removes them when the analysis list is next refreshed.
• LINZ DEM tiles are LERC-compressed. If the GDAL bundled with QGIS cannot read LERC (for example the Flatpak build), the plugin says so and does not fetch.

Re-fetching before analysis

For auto-fetched DEM sources (GSI, Terrarium, Virtual Shizuoka, Nagano, Netherlands, France, New Zealand), the analysis uses the DEM already fetched. Only when the analysis area extends beyond it, the plugin asks whether to fetch it again: OK re-fetches the DEM (and the DSM, if set) and then runs the analysis; No cancels the run. Local DEM/DSM files are analysed on the overlapping part without asking.

TIF / ZIP Layout

Parameter Setting Hints

Slope Stability (Infinite Slope Model)

Computes the slope stability factor of safety (FS) using the infinite slope model.
FS is calculated from slope angle, internal friction angle, cohesion, soil depth, and saturation ratio.
FS < 1.0: collapse risk　/ FS < FS threshold: caution
All parameters describe the properties of the "topsoil down to the hard layer".

How to use the FS threshold (design FS)

FS itself is a physical quantity computed from the four site parameters (φ, c', z, m). The FS threshold is not a physical value — it is how conservative you choose to be, set from the consequence of failure, the permanence of the operation, and the level of uncertainty. It is not a per-site number; site conditions go into the four parameters.

• FS 1.0 is the slope-stability boundary — the state where the forces are in balance.
• FS threshold: how far above 1.0 the "caution" net is cast.

How to read this judgment

The "construction guideline" referenced above is Japan's forest-road construction standard (作設指針). If you are working under a different country's or region's standard, substitute your own local guideline for the same comparison.

Sources
・Shizuoka Prefecture, Economy and Industry Department, Forest Improvement Division, "Guideline for the Construction of Forestry-Dedicated Roads and Forest Operation Roads" (March 2025)
　Shizuoka Prefecture official site: "Construction of Forest Operation Roads"
・Forestry Agency of Japan, "Guideline for the Construction of Forest Operation Roads" (Notice 22-Rinsei-Sei-656, Nov 17, 2010; partially amended by Notice 4-Rinsei-Sei-923, Mar 31, 2023) and its commentary
　Forestry Agency: "Information on the Construction of Forest Operation Roads, etc."

Valley Terrain (TWI)

TWI (Topographic Wetness Index) = ln(A / tan(β)) identifies wet zones.
Higher values indicate valleys/lowlands where water tends to accumulate.

Flow Direction Method (Valley/TWI only)

Choose from 3 methods for how water is routed when computing TWI and valley zones. This choice affects Valley Terrain (TWI) only — flow estimation always uses D8.
It's a trade-off between "how sharp the valley lines are" and "how accurate the wetness distribution is".

D8 (Single Steepest)	D∞ (Tarboton)	MFD (Multiple Flow Direction)
Routes water to a single cell. Produces a sharp, single-line valley, but vertical striping artifacts appear on the TWI surface.	Splits flow between 2 adjacent cells. No striping, and the valley line is reasonably preserved — an intermediate approach.	Distributes flow across all downslope directions. Smoothest wetness distribution. Valleys follow the bottom of the terrain but are thick and diffuse.
Use for: identifying dangerous channels where water concentrates (collapse-prone channels, channels to avoid when routing roads).	Use for: when you want a surface that keeps the line while avoiding striping.	Use for: representing valley terrain accurately as a surface (wetness/hazard surface).

Reference values (same sample DEM as above, 2m, matched valley area — analysis area: 72 ha)

How to read this: A method with a small share and thin lines (D8) points to "dangerous channels." A method with a large share and thick areas (MFD) represents "the valley terrain itself" as a surface.

Because the catchment scale differs by method, the TWI threshold needed to match the same valley area shifts between methods. When switching methods, re-adjust the TWI threshold above (e.g., 5.2 for D8 → around 5.9 for MFD).

Flow Load Estimation (Time-Area Method / Clark Unit Hydrograph)

Estimates the hydraulic load along drainage paths by the time-area method. The upstream area of every cell is split into equal-travel-time bands (isochrones), convolved with a design-storm hyetograph, then routed through one linear reservoir (R = 0.75 × Tc). Three "load" layers are produced:

This is a planning guideline, not an exact hydraulic solution. It assumes a linear, time-invariant response, overland flow on small-to-medium catchments (order of a few km² or less), a single triangular design storm, and D8 flow routing. References: Clark (1945); Chow, Maidment & Mays, Applied Hydrology (1988); USDA NRCS National Engineering Handbook, Part 630.

Output Files

Feature Descriptions

Background processing (progress / stop)

Analysis, DEM/DSM fetching, CS MAP export and WebODM Importer export run in the background, so QGIS stays responsive. During analysis, the Run Analysis button becomes a progress bar showing the percentage, with the current step and the estimated time remaining below it (the estimate is based on the previous run). Stop stops the analysis, and Cancel on the DEM / DSM row stops a fetch. Switching projects or closing the plugin window cancels a running task and removes its partial files.

CS MAP Export

Run Export under Terrain Data > CS Map Export creates a CS-style terrain visualization from the selected DEM. Local DEM files are exported as a whole raster; fetched sources such as VS LP/Grid, GSI DEM, and Terrarium use the currently fetched DEM raster. The output file name is derived from the exported area's center coordinate and size, so exports of different areas never collide or overwrite one another. When Overwrite Same Location is checked (default), re-exporting the same area replaces the previous file for that location; unchecked, each re-export of that same area is kept as a separate file instead. The export is always added to QGIS; when Set As Layer is checked, it is also set as the current Layer Settings > Tile Layer selection.

Example: CS-style terrain output set as the Tile Layer, with Valley Terrain and Flow Estimation overlaid. The tuned colour balance keeps the overlay readable against the terrain.

Slope Stability

Computes the stability factor FS using the infinite slope model. FS < 1.0 indicates collapse risk, FS < 1.5 indicates a caution zone. Calculated from soil parameters (φ, c, γ), terrain slope, and soil depth.

Valley Terrain

Vector data extracted from the DEM, showing catchment terrain and valley lines. Used to check for interference with the drainage network when planning road/skid-trail networks.

Wetness Terrain

Topographic Wetness Index (TWI) raster. Higher values indicate greater water accumulation, poor drainage, weak ground, and higher mud risk.

Flow Load Estimation

Estimates the hydraulic load along drainage paths by the time-area method (Clark unit hydrograph). Produces three layers: Peak load (acute), Mean load (sustained), and Time load (Tc, response speed). Peak and Mean differ by terrain concentration as well as by the storm's peakiness. When DSM is set, the runoff and velocity coefficients vary per cell from canopy height.

Analysis Layer Display (Filter / Buffer)

In the Analysis panel, cycling the Flow button switches Peak load → Mean load → Time load. Filter (Off / Low / Mid) fades out low values to isolate the drainage skeleton — Peak and Mean share one filter setting, Time load has its own. Buffer adds a soft bleed around the flow lines. Peak and Mean share one buffer setting, while Time load has its own independent buffer setting. For an active numbered run the buffer is a temporary display layer. For a Named Export it is stored as a GeoTIFF alongside the analysis files and follows the named layer's visibility.

Integrated Risk

An additive risk index (0–6) combining FS, TWI, and peak load. Extracts areas where multiple hazard factors overlap as "high-priority" polygons.

Named Exports

Select an active numbered run and click Export to Project (first row of Analysis Layer Display, next to the run selector), then enter a unique name. FOL moves the run out of the active sequence into Named Exports and adds the analysis display layers found for that run to QGIS as a persistent group with the same name. The current visibility, opacity, Filter, and Flow Buffer settings become the starting display state. QGIS styles are saved as adjacent .qml files; enabled Flow Buffers are materialised as real GeoTIFFs. The layers keep the same stacking order as the analysis display. The folder also holds the analysis DEM (elev_relief.tif, shown as Elevation Background), so a Named Export stays complete after its source run or the DEM cache is deleted, and its folder can be moved to another project or PC.

Selecting a Named Export already present in the Layers panel connects FOL directly to those real project layers. It does not create a temporary Run <name> copy. Visibility, opacity, Filter, and Flow Buffer changes made through FOL are applied to the named layers and retained. No second export is needed. A Named Export always uses its own display settings: it does not take over the settings of the analysis runs, and switching back to a run restores the run settings used before the Named Export was selected.

Export for WebODM Importer

The Export for WebODM Importer section (Terrain Data tab) packages the loaded terrain data into a ZIP that webodm_importer can load directly. The ZIP is saved in zip/ under the analysis output folder (FOL_YYYYMMDD-all.zip, numbered when exported again on the same day).

• Requirements: webodm_importer is installed, and the DEM source is VS LP/Grid (Shizuoka), AHN4 DTM (Netherlands), IGN LiDAR HD MNT (France) or LINZ DEM (New Zealand). The DSM is optional for every source and is included when DSM Data is set.
• Export is enabled once the requirements are met, and greyed out while the exported ZIP is still valid. Open in WODMI is enabled once a ZIP has been exported and stays enabled until the DTM or DSM changes (a new selection, Clear, or a re-fetch) or the ZIP is deleted. Cancel stops an export in progress.

• For the Netherlands, France and New Zealand, the orthophoto is cut to the DTM extent and grid, and the ZIP also contains DATA_SOURCES.txt: the data list, acquisition dates, sources, licences and attribution. The confirmation and completion dialogs show the same acquisition dates, with a note when the orthophoto and the elevation data were acquired in different years or the orthophoto does not cover the whole area.
• Point clouds are not put into the ZIP. For VS LP/Grid, las_sources.json refers to the LAS files downloaded to create the DSM (vs_lp_ground/), and README_LAS_LINKS.txt lists their download URLs. For the other sources las_sources.json is written with an empty list (it also marks the ZIP as exported from Forestry Operations Lite).
• The orthophoto in the ZIP is a JPEG-compressed GeoTIFF (slightly lossy) to keep the file small.

WebODM Importer Integration

When the current QGIS project contains datasets managed by webodm_importer, their WebODM DTM/DSM assets can be selected directly from the DEM/DSM browser in Forestry Operations Lite.

• Open Browse for DEM or DSM. A WebODM Importer DTM/DSM section appears after the normal local and online data sources.
• The DEM browser lists odm_dem/dtm.tif; the DSM browser lists odm_dem/dsm.tif.
• Datasets exported from Forestry Operations Lite and then imported back through webodm_importer are intentionally excluded, so round-trip FOL outputs do not appear as source data candidates.

Analysis Logic, Limits, and Intent

Forestry Operations Lite does not treat any one layer as the answer. It places several indicators on the same terrain so that their overlap, separation, and location can be examined before forestry work is planned. The values are calculated consistently from the selected data and settings, but they remain indicators of the conditions represented by that data and scenario.

How the current analysis is constructed

• Common terrain: The DEM supplies elevation, slope, terrain form, flow direction, and upstream accumulation. These form the common spatial framework for the analysis.
• Slope Stability: The infinite-slope equation calculates FS from terrain slope and the selected internal friction angle, cohesion, soil depth, and saturation ratio. The same settings are applied across the analysis area as one review scenario; the FS threshold controls how widely caution areas are collected.
• Valley Terrain / TWI: Terrain convergence and slope are combined into a wetness indicator. D8, D-infinity, and MFD describe different ways of distributing upslope contribution, while the TWI threshold and minimum catchment area determine which parts are extracted for review.
• Flow Load: A design storm is combined with D8 travel time and a time-area / Clark-style linear-storage calculation. Peak, Mean, and Time load show where the selected scenario concentrates quickly, persists, or responds slowly. When a DSM is available, canopy height varies the runoff and velocity coefficients at valid cells; the editable fallback values are used elsewhere. This is a simplified numerical runoff scenario, not an implementation of HEC-HMS or a measured hydrograph.
• Overall Risk Index: FS, TWI, and Peak load scores are added to reveal places where indicators overlap. The sum is a review priority, not a probability of failure and not proof of a causal relationship between the component layers.

Limits to keep visible

• The soil and saturation settings express one condition across the area. They do not map every local change in soil layer, root system, compaction, groundwater, or recent rainfall.
• DEM and DSM represent their acquisition date, resolution, classification, and processing. Canopy height does not by itself distinguish canopy density, understory, soil properties, or non-forest objects contained in a DSM.
• Grid resolution, NoData, pits and flats, flow-direction method, thresholds, and design rainfall all affect the pattern. Values with physical units are still model responses under the selected conditions, not observations of the site.
• The outputs do not replace field judgment, geotechnical or hydraulic investigation, structure design, or the applicable forestry-road and drainage standards. They show where those questions deserve attention.

Why FOL uses indicators

Calling the outputs indicators is not a denial of analysis. More accurate terrain data, better parameters, and improved methods can bring an analysis closer to actual conditions. That improvement can continue, but data and computation do not become the field itself. The same applies to larger or institutionally published models: more components provide a different and often more detailed approximation, not an automatic claim to truth.

Field observation is therefore not used merely to certify whether the map is correct. The map helps decide where to look; the site supplies context that the data does not hold; and the practitioner updates the work plan from both. Slope Stability, Valley Terrain, Flow Load, background terrain, and existing roads should be read together when considering road alignment, harvesting and extraction, drainage, crossings, or areas to retain and protect.