Have a personal or library account? Click to login
The Potential of Landscape Metrics for Estimating Forest Fire Risk in Poland Cover

The Potential of Landscape Metrics for Estimating Forest Fire Risk in Poland

Quaestiones Geographicae
Volume 45 (2026): Issue 1 (March 2026)
By: Aleksandra Kolanek,  Mariusz Szymanowski and  Tomasz Niedzielski  
Open Access
|Mar 2026

Figures & Tables

Fig. 1.

Differences between the structure of forest stands in general forest areas (outer ring) and in stands where fires broke out (inner ring); data from the entire territory of Poland, 2007–2017. Units are given in percentages.

Fig. 2.

Forest fires occurred in 2007–2017 in Poland (A) and in 2015 in Lubelskie Voivodeship (B).

Fig. 3.

Examples of patch distribution in buffer zones belonging to individual classes for the three analysed stand features. Data source: Forest Data Bank (2018). Each forest subarea has specific attributes (top row). Adjacent subareas with the same attributes were merged to forest patches (bottom row).

U Mann-Whitney test of significance of differences between test (A1) and control (B and C) groups for a class level in the Lubelskie province (with significance level of 0_05); * _ statistically significant difference; – – non-significance; × – no fulfillment of the Ljung-Box test assumptions (with significance level of 0_01)_

forest featureclassmetricmean valuessignificance of differencedirection of difference (group A1 compared to other(s))
group A1group Bgroup CA1 vs BA1 vs C
age class0–20NumP10.749.6210.08*–↑
MPS15,843.9518,852.8422,300.089–*↓
MPE598.63643.17699.68–*↓
21–40NumP14.5811.70×*×↑
MPS16,007.82×23,208.13×*↓
PSSD18,196.26×24,523.71×*↓
MPE572.97×699.26×*↓
41–60MedPS10,261.8816,281.8122,505.81**↓
MPE716.95869.33838.83**↓
MPAR0.160.140.13–*↑
above 100MPS48,430.8982,508.2090,334.95**↓
MedPS34,246.0159,932.5564,983.52**↓
PSSD36,922.7263,068.5174,370.46–*↓
TE3,902.175,077.765,848.49–*↓
ED0.00120.00160.0019–*↓
MPE940.871,313.201,328.98**↓
trophic typeconiferousMPAR0.130.110.23*–↑
mixed (coniferous dominance)MPS50,221.7997,672.7276,815.13**↓
PSSD76,474.80135,536.72109,243.81**↓
TE13,611.1714,224.8014,768.00*–↓
ED0.00430.00450.0047*–↓
MPE965.201,351.901,202.55**↓
AWMSI1.87672.02932.0349*–↓
mixed (deciduous dominance)MPS65,612.30125,744.80129,248.48**↓
TE8,056.0111,467.2210,461.48**↓
ED0.00260.00360.0033**↓
MPE1,044.971,397.511,425.78**↓
AWMSI1.741.911.88**↓
trophic typedeciduousMedPS32,883.01×151,401.46×*↓
PSCoV74.4698.49104.39**↓
TE3,752.868,102.727,812.92**↓
ED0.00120.00260.0025**↓
MPE901.561,615.181,786.23**↓
MPFD1.411.381.39–*↑
AWMPFD1.361.341.33**↑
riparianPSSD1,018.4428,118.008,191.14*–↓
humidity typelowland/upland boggyMPS32,655.4645,032.8847,329.74**↓
MedPS26,755.3426,933.4933,966.39**↓
PSSD27,515.5448,934.7234,810.27**↓
TE3,565.165,044.874,777.90*–↓
ED0.00110.00160.0015*–↓
MPE795.191,008.651,025.47**↓
AWMPFD1.371.361.36–*↑
lowland/upland humidMPS49,006.5464,442.6566,420.40**↓
MedPS18,806.6123,624.0336,780.94**↓
MSI1.701.761.73*–↓
AWMSI1.862.051.94**↓
lowland/upland freshNumP20.3115.40×*×↑
MPS11,4533.22277,913.46278,704.02**↓
PSSD18,3262.98257,122.58×*×↓
MPE1,441.721,929.301,977.48**↓
AWMPFD1.321.31×*×↑
lowland dryMPS59,539.8219,212.0922,153.63**↑
MedPS54,795.9016,524.3914,166.96–*↑
MPE1229.21633.41598.16**↑
MPAR0.04100.05090.0591–*↓

Landscape metrics used in this study_

patches featureMetric name and acronymexplanation
number of forest patchesNumber of patches (NumP)
  • –

    total number of patches in the landscape (landscape level)

  • –

    number of patches for each class (class level)

size of forest patchesMean patch size (MPS)
  • –

    average patch size of all patches (landscape level)

  • –

    average patch size of each class patches (class level)

Median patch size (MedPS)
  • –

    median patch size of all patches (landscape level)

  • –

    median patch size of each class patches (class level)

Patch Size Standard Deviation (PSSD)
  • –

    standard deviation of all patch size (landscape level)

  • –

    standard deviation of patch size within class (class level)

Patch Size Coefficient of Variance (PSCoV)
  • –

    relative variability of all patch sizes within a landscape (landscape level)

  • –

    relative variability of patch size within class (class level)

edges of forest patchesEdge density (ED)The amount of patch edges relative to the patch area:
  • –

    edge density of all classes (landscape level)

  • –

    edge density of each class (class level)

Mean patch edge (MPE)The average amount of edge per patch
  • –

    for all patches (landscape level)

  • –

    for each class patches (class level)

forest patches complexityMean shape index (MSI)Takes values equal to or greater than 1, when patches are circular, then MSI = 1
  • –

    for all patches (landscape level)

  • –

    for each class patches (class level)

Area-weighted mean shape index (AWMSI)MSI related to patch size
  • –

    for all patches (landscape level)

  • –

    for each class patches (class level)

Mean perimeter-area ratio (MPAR)Perimeter of patches related to patch area:
  • –

    for all patches (landscape level)

  • –

    for each class patches (class level)

U Mann-Whitney test of significance of differences between test (A1) and control (B and C) groups for a landscape level in the Lubelskie province (with significance level of 0_05); * – statistically significant difference; – – non-significance; × – no fulfillment of the Ljung-Box test assumptions (with significance level of 0_01)_

forest featuremetricmean valuessignificance of differencedirection of difference (group A1 compared to other(s))
group A1group Bgroup CA1 vs BA1 vs C
age classMedPS8,728.0611,059.75×*×↓
MPE1,003.291,061.111,065.97**↓
AWMSI3.042.87×*×↑
trophic typeNumP56.3043.60×*×↑
MPS98,897.47133,178.45142,389.22**↓
PSCoV355.83319.81×*×↑
MPE1,299.431,470.541,507.99**↓
AWMSI3.062.91×*×↑
humidity typeNumP49.2638.2939.93–*↑
MPS107,686.13144,325.46159,496.67–*↓
PSCoV344.65314.08×*×↑
MPE1,386.211,542.721,575.72**↓
AWMSI3.132.972.81**↑

Share of forest where fires occured in 2015 in relation to share of total forest area in Lubelskie Voivodeship in 2015_ Prepared based on the Forest Data Bank (2018) and the National Forest Inventory (2011–2015) (https://www_bdl_lasy_gov_pl/portal/wisl-en)_

Tree stand characteristicshare [%] of total forest areashare [%] of forests, where fires occured
trophic typeconiferous17.848.13
mixed (conifer–dominated)28.027.5
mixed (deciduous–dominated)27.920.0
deciduous26.34.0
humidity typefresh80.086.3
humid14.89.4
other5.24.3
age class (in years)0–209.214.9
21–4015.613.13
41–6026.624.7
61–8025.936.4
81–10013.28.4
more than 1005.01.9
DOI: https://doi.org/10.14746/quageo-2026-0010 | Journal eISSN: 2081-6383 | Journal ISSN: 2082-2103
Journal RSS Feed
Language: English
Page range: 155 - 170
Submitted on: Jun 10, 2025
|
Published on: Mar 12, 2026
Published by: Adam Mickiewicz University
In partnership with: Paradigm Publishing Services
Publication frequency: 4 issues per year
Keywords:
forest fires,
landscape metrics,
fire risk,
forest management
Related subjects:
Geosciences,
Geography

© 2026 Aleksandra Kolanek, Mariusz Szymanowski, Tomasz Niedzielski, published by Adam Mickiewicz University
This work is licensed under the Creative Commons Attribution 4.0 License.

Previous articleVolume 45 (2026): Issue 1 (March 2026)Next article