CCFPI research

The completed MSc Forestry research model, presented faithfully and kept separate from the application's demonstration configuration.

Research record

Assessing Ireland's suitability for continuous cover forestry: Development of the Continuous Cover Forestry Potential Index (CCFPI) for national application and Sitka spruce transformation

Author
Jay Gilbert
Position
MSc Forestry researcher · Independent Researcher and Founder, BRANCHX LIMITED
Status
Completed MSc Forestry research · 72 pages
Focus
National application and Sitka spruce transformation
ORCID iD 0009-0003-7255-5187

ORCID iD is a persistent identifier for the researcher. It is not a validation, accreditation or endorsement of the research.

Completed MSc Forestry research. It is not a draft, and it is not peer-reviewed journal research. No journal publication is claimed.

The full thesis file is not attached to this build. No download is offered because no thesis asset exists in this repository.

The original MSc research model is the seven-factor Fuzzy Weighted Overlay shown on this page. The CCFPI configuration used in the application's lifecycle demonstration is a simplified operational/demo implementation with a different factor set and different weights. Those demonstration weights are not the original MSc research weights.

Purpose

GIS-based screening for continuous cover potential

GIS-based decision support for screening Continuous Cover Forestry potential, developed for national application in Ireland with a specific focus on Sitka spruce transformation.

GIS-MCDA

Spatial multi-criteria decision analysis over national terrain, soil and wind datasets, combining standardised factor surfaces into a single potential index.

Fuzzy membership

Each factor surface is transformed into a continuous 0–1 membership so partial suitability is preserved rather than forced into hard classes.

Fuzzy Overlay (FO)

Unweighted fuzzy combination of the factor memberships, used as the comparison method in the validation exercise.

Fuzzy Weighted Overlay (FWO)

The adopted CCFPI method: each fuzzy membership is multiplied by its research weight and summed to produce the index.

ArcGIS Pro

Preprocessing, fuzzy membership generation and the Raster Calculator overlay were carried out in ArcGIS Pro.

Method

How the index was produced

The research workflow, in the order it was carried out in ArcGIS Pro.

  1. National input datasets
  2. Preprocess and align to analysis grid
  3. Fuzzy membership per factor
  4. Research weights
  5. Fuzzy Weighted Overlay
  6. Suitability classes
  7. Sensitivity and Wicklow concordance
Original research model

Seven factors and their research weights

These are the factors and weights of the original MSc Fuzzy Weighted Overlay equation.

  • DEM / elevation

    25%

    FuzzyMe_dem · Elevation sets the broad climatic and exposure envelope for continuous cover structures.

  • Slope

    21%

    FuzzyMe_Slope · Slope governs operability, extraction and the practicality of repeated selective intervention.

  • Wind speed

    17%

    FuzzyMe_ws50 · Wind loading is the dominant risk to the residual stand once the canopy is opened.

  • Soil texture

    14%

    FuzzyMe_soil_texture · Texture influences rooting, stability and the response of natural regeneration.

  • Soil depth

    10%

    FuzzyMe_soil_depth · Rooting depth conditions anchorage and the persistence of a retained overstorey.

  • Soil hydrology

    7%

    FuzzyMe_soil_hydro · Drainage status affects both stability and regeneration establishment.

  • Aspect

    3%

    FuzzyMe_Aspect · Aspect contributes a small directional adjustment for exposure and light.

CCFPI = (FuzzyMe_dem × 25) + (FuzzyMe_Slope × 21) + (FuzzyMe_ws50 × 17) + (FuzzyMe_soil_texture × 14) + (FuzzyMe_soil_depth × 10) + (FuzzyMe_soil_hydro × 7) + (FuzzyMe_Aspect × 3)

The equation as entered in the ArcGIS Pro Raster Calculator, reproduced from the MSc. The seven stated weights sum to 97 rather than 100; the values are shown as published rather than rescaled.

Reported results

National and Sitka spruce outcomes reported in the MSc

Figures reported in the completed MSc research. No national raster asset is attached to this build, so results are shown as research-result visualisations rather than as a map.

Ireland — national
41.05%

The MSc reports that 41.05% of Ireland's land area is potentially Suitable or Very Suitable for Continuous Cover Forestry (13.83% Very Suitable plus 27.22% Suitable).

Very Suitable 13.83%Suitable 27.22%
Sitka spruce transformation
22.1%

For Sitka spruce, the MSc reports 22.1% of the analysed area as Suitable or Very Suitable. The class values shown are the explicit source chart/table values.

  • Very Unsuitable7.04%
  • Unsuitable53.44%
  • Intermediate17.42%
  • Suitable16.21%
  • Very Suitable5.89%

Research note: the thesis text elsewhere states Intermediate as 15.47%, which differs from the 17.42% in the source chart. The discrepancy is shown rather than silently reconciled; source-version reconciliation is required before any journal submission.

Sensitivity analysis

How much each weight moves the result

Factor weights were adjusted by ±10%, ±20% and ±30% and the resulting change in land-area class allocation was recorded.

  • Slope ±30%

    18% of land area shifted suitability class

  • Wind speed ±20%

    15% of land area shifted suitability class

  • Soil texture ±20%

    10% of land area shifted suitability class

  • Soil depth ±20%

    10% of land area shifted suitability class

  • Aspect ±10% to ±30%

    Less than 5% of land area shifted class

  • Soil hydrology ±10% to ±30%

    Less than 5% of land area shifted class

The research concluded that slope and wind speed were the most influential factors in the index.

Initial validation

Concordance against known CCF sites in County Wicklow

An initial validation exercise compared model output against known Continuous Cover Forestry sites in County Wicklow, including Sitka spruce plots.

FWO score range
46.17–95.01

Across most Wicklow CCF plots.

Reported FWO score
89.35

Average ‘accuracy score’ as reported in the MSc.

Reported FO score
57.26

Unweighted Fuzzy Overlay comparison.

The MSc reports an average 'accuracy score' of 89.35 for Fuzzy Weighted Overlay against 57.26 for unweighted Fuzzy Overlay. These figures are reported in the MSc and should be read in the context of that thesis's validation design and metric definition. They are not independently validated predictive accuracy, and nothing here is certified.

Limitations

What the research does not claim

CCFPI is a screening and decision-support index. It is not a substitute for professional forestry judgement or site assessment, and it carries no certification, compliance or regulatory-approval status.

  • Source datasets vary in resolution, and local microclimate effects are not resolved.
  • Resampling coarser source data onto a 10 m analysis grid does not create true 10 m source information; the analysis grid is finer than the evidence behind it.
  • Long-term uncertainty from climate change, pest and disease pressure and policy change is not modelled.
  • Species responses differ; results for Sitka spruce should not be transferred to other species without new work.
Provenance and versioning

Which model, which build, which claim

Research model
MSc research model (as submitted)
Model identifier
ccfpi-msc-research
Software implementation
The current software implementation of CCFPI has not been independently validated or reproduced against the original research outputs.
Research validation
The MSc itself included an initial validation exercise against known Continuous Cover Forestry sites in County Wicklow.
Bridge into MyForestPass

Apply the model to a forest

Inside MyForestPass, a model run is recorded as evidence: the model version, the input surfaces and their provenance, the resulting score and the reviewer who accepted or rejected it. The research model above is one example of a model whose output can be carried in that evidence record.

  1. 01Register the mapped forest extent as the persistent unit of continuity.
  2. 02Attach the input surfaces with their source, resolution and date.
  3. 03Record the model run and its version alongside the score it produced.
  4. 04Record the professional review as a separate assertion, not an overwrite.
Software implementation methodology Evidence fitness