=Paper=
{{Paper
|id=Vol-2534/78_short_paper
|storemode=property
|title=GIS mapping of forest paludified landscape in the Great Vasuygan Mire marginal area (Western Siberia)
|pdfUrl=https://ceur-ws.org/Vol-2534/78_short_paper.pdf
|volume=Vol-2534
|authors=Anna A. Sinyutkina
}}
==GIS mapping of forest paludified landscape in the Great Vasuygan Mire marginal area (Western Siberia)==
https://ceur-ws.org/Vol-2534/78_short_paper.pdf
GIS mapping of forest paludified landscape in the Great Vasuygan
Mire marginal area (Western Siberia)
Anna A. Sinyutkina
Siberian Research Institute of Agricultural and Peat – branch of Siberian Federal Scientific Centre of Agro-
Biotechnologies, Tomsk, Russian Federation, ankalaeva@yandex.ru
Abstract. The aim of research is to estimate forest state within the Great Vasuygan mire
marginal part using satellite and field data. Specifically, the objective of this study were
to: 1) conduct field landscape studies and assessment of paludification intense to create
training samples and asses the accuracy of satellite images interpretation; 2) distinguish
paludification landscapes in mix forest surroundings raised bog massif by a semi-
automatic landscape classification method and assessment mire influence zone; 3) create
the classification of the raised bog massif spatial structure. The average length of mire
influence zone is 3.9 km with fluctuations in the range of 0.6–8.4 km. Spatial structure is
indicator of the development stage and the slope of the raised bog and, as a consequence,
the intensity of surface runoff from mire into the surrounding areas. The length of mire
influence zone differs between the selected classes of spatial structures and is mainly
determined by the distribution of complex microlandscapes, the presence of which
reflects the final stage of raised bog massif development.
Keywords: Landsat; mire classification; swamp forest; satellite image interpretation;
GPR survey; peat deposit
1 Introduction
Mires are valuable natural resources that provide many environmental services to flora and fauna. Recently,
wetlands have become a popular topic in climate change discussions as they include 12% of the global carbon pool
[1]. Climate change and rapid land use change have turned mires into carbon source ecosystems. Peat mining,
drainage, agriculture and potential negative feedback with warming environment are releasing the carbon stored in
peat, and adding to atmosphere carbon dioxide [2]. Meanwhile, vast pristine wetland areas remain on the territory of
Western Siberia and, in particular within the Great Vasyugan Mire, and the process of waterlogging continues [3].
The marginal parts of large raised bog massifs are most exposed to paludification. The process of paludification is
mainly due to an increase in the water table level, leading to anaerobic conditions that in turn reduce microbiological
activity and limit mineralization and nutrient uptake by plants [4]. The ambiguity of the paludification trend estimates
within the mires marginal parts determines the need to obtain new data on the current carbon stocks here [5]. GIS
mapping techniques can help generate accurate map of wetlands and get the spatial estimation paludification area.
These maps can be used in climate models to assess the sensitivity and feedback to future climate change [2].
It is known that the length of the influence zone of the mire is mainly determined by geological and
geomorphological factors, among which are the slope, horizontal and vertical surface defragmentation. The intensity
of the paludification process development can be influenced by the change of the surface slope even by tenths of a
degree. The marginal part of the Great Vasyugan Mire belongs to the territories with low values of the surface slope.
Therefore, the dependence of the mire impact zone length and the surface slope is not always determined here by the
medium-scale mapping due to the lack of detail and accuracy of topographic maps and SRTM data. The assessment
of the relief morphometric parameters influence on the paludification process becomes possible only by detailed field
studies in key sites [6]. Therefore, available indicators of the paludification process is necessary to search for the
modeling of potential paludification zones which are convenient to use and allow to assess the hydromorphic
transformation on large areas. One of these indicators is the characteristic of the spatial structures of the raised bog
massifs, reflecting the stage of development of the mire, the shape of its surface and, as a consequence, the intensity
of the surface flow to the surrounding areas.
The aim of research is to estimate forest state within the Great Vasuygan mire marginal part using satellite and
field data. Specifically, the objective of this study were to: 1) conduct field landscape studies and assessment of
paludification intense to create training samples and asses the accuracy of satellite images interpretation; 2)
distinguish paludification landscapes in mix forest surroundings raised bog massif by a semi-automatic landscape
Copyright © 2019 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0
International (CC BY 4.0).
classification method and assessment mire influence zone; 3) create the classification of the raised bog massif spatial
structure.
2 Methods and data
2.1 Study area
The objects of study are the North-Eastern spurs of the Great Vasyugan Mire – Iksa and Bakchar raised bog
massif (figure 1). The study area is located in the southeast West Siberian plain within Middle Ob River watershed.
The territory belongs biogeographically to the south taiga zone. The fluviolacustrine loams and clays represent a
quaternary deposit. The quaternary deposits thickness on the interfluve of the Bakchar River and the Iksa River
reaches 40-60 m [7]. The climate is continental with long, cold winters and short, hot summers; the average annual
temperature is 0.23°C. The annual amount of precipitation is 473 mm according to the meteorological station near the
Bakchar village. The average annual evapotranspiration reaches 332 mm. Positive atmospheric water balances, flat
relief, and weak drainage by rivers allows the formation and sustainable evolution of mires [8]. Large mire massifs
are widely distributed within the study area. The study area includes typical Western Siberia pine-shrub-sphagnum,
pine-shrub sedge-sphagnum ombrotrophic mires, and swamp forest with birch, aspen, Siberian cedar, and spruce in
the margin part of the mire.
Figure 1. Study area (Landsat 8, July 17 2018)
2.2 Field study
The GPR surveys, which covered a total distance of 2.3 kilometers, were conducted in March 2017 and 2018. We
employed a GPR system “OKO-2” (“Logical systems”, Russia) with 250 MHz shielded antenna and displacement
sensor. We placed marks as vertical lines on radiograms every 50-100 m during the GPR surveys to collect peat core
and to binding high-altitude data in summer. The step size between the marks was determined by the peat deposits
heterogeneity. The marks, the beginning and end of the GPR transect were located with GPS (accuracy: 5 m).
In accordance with the marks on the GPR profile in the summer of 2018 made contact sensing of peat deposits on
33 points, including definition of peat deposits depth, the allocation of stratigraphic layers by the types and degree of
peat decomposition by the visual method [9, 10]. Geobotanical study are made, the height of the trees was determined
using a laser rangefinder Nikon Forestry Pro. The mire water level is determined on 12 sites. Determination of the
habitats soil moisture stages was carried out in accordance with [11] taking into account the specific weight of the
plant species projective cover [5, 6]. The paludification intense assessment was carried out using the indicators: the
stage of habitat moisture; the depth of peat deposits and the depth of the fibric peat; the groundwater level.
2.2 GIS mapping and spatial structure classification
The paludification zone definition was conducted on the basis of Landsat 8 data classification using Semi-
Automatic classification plugin. The training sample was compiled using the data of field landscape studies and
included 4 classes of sites indicating the development of paludification processes within forest: 1) swamp forest areas
characterized by mass felling of trees; 2) pine-shrub grass-sphagnum bogs outside the raised bog massif main
contour, 3) cedar-birch swamp forest; 4) pine-birch swamp forest. Assessment of the mire influence zone length is
determined by the method of landscape profiling. We laid 40 profiles crossing the border of raised bog massifs
parallel to the drain lines. We measured. the bog massif length, the distance from the watershed to the marginal part
of the mire, the extent of the mire influence zone, the surface slope within the influence zone, a set and the length of
the main microlandscapes (hummock-hollow and hummock-pool complexes, pine-dwarf shrubs-sphagnum bogs,
grass-moss mires). The length of mire influence zone was measured in accordance with the distribution of plots
allocated by the results of Landsat image interpretation on each profile. The surface slope map is drawed by the
method of raster interpolation Topo to Raster based on the elevation of topographic maps with correction on the lines
of the hydrographic network.
We used the method of cluster analysis to classify spatial structures of raised bog massifs. The comparison of
values paludification zone length differences between the classes was conducted using tests of Mann-Whitney and
Kruskal-Wallis. The dependence of the paludification zone length, the characteristics of the spatial structure and the
surface slope was determined using the principal components method.
3 Result and discussion
3.1 Paludification landscape within the mire influence zone
Swamp forests with a predominance of Betula pubescens are widespread within the key area. Subdominants of the
tree layer are Pinus sibirica, Pinus sylvestris, Picea obovata, Populus tremula. The tree layer height varies from 15 to
22 m with a cover of 20-60 %. Typical representatives of the shrub layer are Ribes nigrum, Rosa acicularis, Sorbus
sibirica, Rubus idaeus. The layer projective cover usually does not exceed 5-10 %. Various species of sedges (Carex
acuta, C. Elongata, etc.), horsetails (Equisetum palustre, E. Sylvaticum, etc.), reed grasses (Calamagrostis canescens,
C. neglecta, etc.), ferns (Gymnocarpium dryopteris, Athyrium filix-femina, etc.) are widely distributed in the grass
layer. Bryidae occupying a fallen rotten trunks and tree trunks hummock dominate in the moss layer often [6]. The
stage of habitat moisture is characterized as wet-forest, wet-forest and wet-forest [11].
Shallow peat soils with a peat layer depth 0,7–1 m and peat-gley soil with a peat layer less 0,7 dominate within the
mire influence zone. The peat deposits depth within the key area reaches 1.75 m, the average value is 0.4 m. Sapric
wood and wood-grass peat with a high degree of decomposition dominate in the stratigraphy of swamp forest peat
deposits.
3.2 GIS mapping result of raised bog massif and it influence zone
Spatial structure of large raised bog massif within the Great Vasyugan Mire spruces is characterized by a change
of plant communities in the following sequence: 1) pine grass-sphagnum bogs, stretching a narrow strip along the
mire border, 2) the pine-shrub-sphagnum bog, occupying a large area within the raised bog massifs slopes, 3) ridge-
hollows and ridge-pool complex pine-shrub grass-sphagnum bogs in the central part of the raised bog massifs. The
large raised bog massif northern parts and smaller raised bog massif are characterized by a lack of complex bogs and
a simpler spatial structure. The averages width of the raised bog massifs in parallel to the direction of flow is around
11 km, in some areas up to 24 km. The surface slope within the mires ranges 0-0,2°, but for the most part it does not
exceed 0.05°. The surface slope does not exceed 0.3° and an average of about 0.1° within the watershed plains
occupied by forests and swamp forest (figure 2).
According to principle component analysis the mire influence zone length is largely determined by the raised bog
massif width (0.92), a length of ridge-hollow (0,64) and ridge-pool (0,63) complex bog. The surface slope of the areas
adjacent to the mire have less influence (0.30) due to the low variation of the values with almost flat relief within the
area under study (figure 3).
Figure 2. Surface slope map of key site Figure 3. Principle component analysis result: Zone –
mire influence zone; Width2 – distance from the
watershed to the marginal part of the mire; Slope – surface
slope; microlandscapes length: M-g – moss-grass; Tree –
wood and wood-moss; gmk – ridge-hollows complex; gok
– ridge-pool complex.
3.3 Classification of raised bog spatial structure
The raised bog massif spatial structures are combined to four classes as a result cluster analysis of bog
microlandscapes length data array.
The first class of spatial structures is characterized by a wide spread of complex bogs, including ridge-pool
microlandscapes. The length of ridge-pool microlandscapes reaches its maximum value within the considered area of
15 km, with an average value of 8.7 km (figure 4, 6). Unlike the first, the second class is characterized by a large
distribution of ridge-hollow complex, with absence ridge-pool complex. The raised bog massif marginal parts are
occupied by wood and wood moss microlandscapes. The mire influence zone length values are maximum in the first
and second spatial structures class within the key area. The length reach to 8.4 km from the border of the raised bog
massif with an average value of 4.6 km.
Figure 4. Mire influence zone with first class spatial structure (Landsat 8, July 17 2018)
The third spatial structure class is characterized by a wide distribution of the grass-moss microlandscapes. The
average length of ridge-hollow complex does not exceed 1 km, ridge-pool complexes are absent in most cases. The
average length of the mire influence zone is 3.5 km. The fourth spatial structures class is characterized by a lack of a
complex and wide distribution of wood and wood moss micro-landscapes. The values of mire influence zone are
minimum length in the fourth class within the site under study which average made 1,7 km (figure 5,6).
Figure 5. Mire influence zone with fourth class spatial structure (Landsat 8, July 17 2018)
Figure 6. Comparison of the mire influence zone length of the spatial structure classes (I–IV – class numbers)
Thus, the study result is assessment of the mire influence zone and forest state within the marginal part of the
Great Vasyugan Mire North-Eastern spurs. The average zone length is 3.9 km with fluctuations in the range of 0.6–
8.4 km. The dependence of the mire influence zone length from the characteristics of the spatial structure is
statistically confirmed. Spatial structure is indicator of the development stage and the slope of the raised bog surface
and, as a consequence, the intensity of surface runoff from mire into the surrounding areas. The length of mire
influence zone differs between the selected classes of spatial structures and is mainly determined by the distribution
of complex microlandscapes, the presence of which reflects the final stage of development of the raised bog massif.
4 Conclusion
The use of modern research methods based on the analysis of satellite data and GIS mapping allowed to obtain
qualitatively new scientific data on the spatial differentiation of adjacent areas to the mires and to assess the length of
the influence zone of large raised bog massif. The developed algorithms for modeling the hydromorphic
transformation of ecosystems will be universal for the taiga zone of Western Siberia and can be used to assess the
paludification processes in large areas.
Acknowledgments
This study was funded by RFBR, research project no. 18-35-00387.
References
[1] Amani M., Salehi B., Mahdavi S., Brisco B. Spectral analysis of wetland using multi-sourse optical satellite
imagery // ISPRS Journal of Photogrammetry and Remote Sensing. 2018. Vol.144. P. 119-136.
[2] Minasny B., Berglund Ö., Connolly J., Hedley C., de Vries F., Gimona A., Kempen B., Kidd D., Lilja H.,
Malone B., McBratney A., Roudier P., O'Rourke S, Rudiyanto, Padarian J, Poggio L., Caten A., Thompson D.,
Tuve C., irastuti Widyatmanti W. Digital mapping of peatlands – A critical review // Earth-Science Reviews.
2019. Vol. 196. 102870
[3] Babikov B.V., Kobak K.I. Absorption of Atmospheric Carbon Dioxide by the Wetland Ecosystems of Russia in
Holocene. The Problems of Рaludification, Bulletin of Higher Educational Institutions. Forestry journal. 2016.
№ 1(349). P. 6–36.
[4] Mansuy N., Valeria O., Laamrani A., Fenton N., Guindon L., Bergeron Y., Beaudoin A., Légaré S. Digital
mapping of paludification in soils under black spruce forests of eastern Canada // Geoderma Regional. 2018.
Vol. 15. e00194
[5] Sinyutkina A.A. Investigating the peat deposits of the Great Vasyugan Mire margin using ground-penetrating
radar // IOP Conference Series: Earth and Environmental Science. 2018. Vol. 211 (1). 012066
[6] Sinyutkina A.A., Gashkova L.P. Assessment of the status of the Bakchar bog marginal part of the Great
Vasyugan mire // The Journal of Soils and Environment. 2018. № 1(4). 243 –255.
[7] Richter G. D. Relief and geological structure Western Siberia. Moscow: Publishing House of the USSR
Academy of Sciences),1963. P. 22-68.
[8] Evseeva N. S., Sinyutkina A.A. and Kharanzhevskaya Yu. A. et al. Landscape of the mires in Tomsk Region.
Tomsk: Publishing house of NTL, 2012. P. 400.
[9] Tyuremnov S. N. Peatlands. Moscow: Nedra, 1976. P. 488.
[10] von Post L. The Swedish Revolution of the Geological Survey, and some of its findings // The Swedish Mosque
Society Association's Journal. 1922. №1. P. 1-27.
[11] Tsyganov D.N. Phytoindication of ecological regimes in the subzone of coniferous-deciduous forests. Moscow,
Nauka Pbs, 1983. P. 197.