Quantifying Carbon Reserves in Urban-Encroached Peat Soils of Equatorial Borneo

Tropical peatlands store an estimated 88.6 Pg of carbon (range 81.7–91.9 Pg), of which approximately 68.5 Pg (77%) is concentrated in Southeast Asia—a disproportionate share relative to their areal extent [1-3]. Intact peatlands function as persistent net carbon sinks [4, 5] but once drained or burned rapidly become major greenhouse gas sources [6], with current CO₂ emissions from drained Southeast Asian peatlands reaching approximately 355 Mt CO₂·yr⁻¹ [7, 8]. Indonesia holds approximately 21 million hectares—roughly 80% of the Southeast Asian total—with estimated carbon stocks of 13.6–40.5 Gt C across Sumatra, Kalimantan, and Papua [9-11]; accurate sub-national peat carbon inventories from sites undergoing active urban development remain scarce, despite their importance for understanding carbon dynamics under land-use pressure [12]. Borneo (Kalimantan) hosts some of the world’s most carbon-rich peat swamp forests [13]; in West Kalimantan, Pontianak City—directly underlain by tropical peat of variable depth—faces continuous residential housing expansion that disturbs peat stratigraphy, alters drainage, and accelerates oxidative carbon loss [14, 15], yet district-level carbon stock data remain sparse. Throughout this paper, “urban encroachment” denotes the expansion of residential housing into the peatland margin—identified from land-use records as the dominant disturbance driver in both study districts—rather than industrial, agricultural, or drainage-canal development. “Carbon reserves,” as used in the title, likewise denotes in-situ subsurface carbon stocks derived from volumetric calculation of peat depth and carbon density, rather than economically extractable reserves in the resource-economics sense.

Carbon stock quantification in tropical peat relies on three primary parameters: bulk density, organic carbon concentration via loss on ignition (LOI) with the Van Bemmelen conversion factor, and deposit thickness [16-20]. In urban peat environments where drainage-induced compaction causes bulk density to deviate from pristine references, a specimen-level check of material identity against direct elemental measurements is a useful complement [21, 22]. Scanning Electron Microscopy coupled with Energy-Dispersive X-Ray Spectroscopy (SEM-EDX) addresses this by revealing fibrous peat’s dual-porosity architecture—macropores between fibre elements and micropores within the fibre matrix—and providing quantitative elemental composition that confirms genuine, low-mineral-content peat for a representative specimen, without itself validating LOI-derived organic fractions at district scale [23-26]. Translating point-based field measurements of peat depth, groundwater level, and carbon density into continuous district-scale surfaces requires a spatial interpolation method suited to the sampling density and objectives of the study. Prior studies in Indonesian peatlands have applied fitted geostatistical interpolators—ordinary kriging outperforming inverse-distance weighting (IDW) and machine-learning alternatives for peat thickness and carbon-stock mapping in one case [27]—while a second study mapped peat thickness and groundwater level directly along geophysical survey transects, without an intervening spatial interpolation step [28], and a third applied kriging-based peat-thickness modelling [29]. This study instead adopts a Thiessen (Voronoi) proximity-polygon zonation for all three mapped layers (peat depth, groundwater level, and carbon density), detailed with its supporting citation in Section 2.4: each sample point defines a discrete zone of influence extending to the midpoint between it and its nearest neighbours, with the point’s measured value applied uniformly across that zone. This approach was selected over geostatistical alternatives such as kriging or IDW because it is a deterministic method that requires no parameters to be fitted or tuned to the data—unlike kriging’s variogram model or IDW’s power exponent and search radius. Any analyst working from the same point dataset will reproduce an identical Thiessen zone map, with no model-selection step introducing additional, unreported uncertainty.

Despite substantial literature on Indonesian peat carbon stocks at regional and provincial scales, to the best of our knowledge, a critical research gap persists: integrated, district-level assessments that combine SEM-EDX microstructural characterisation with field-calibrated, GIS-based spatial carbon mapping for rapidly urbanising coastal peat environments. Existing studies have predominantly focused on natural peat swamp forests and agricultural conversion areas in Riau Province, Central Kalimantan, and Papua, with insufficient attention to urban peat settings, where ongoing land development continually modifies carbon stock dynamics [30]. The North Pontianak (Pontianak Utara) and Southeast Pontianak (Pontianak Tenggara) districts represent an underrepresented yet scientifically significant study domain at the active frontier of urban expansion over the tropical peatlands of Borneo. This study is distinct from the authors’ related recent work on the same delta: Parabi et al. [14] addressed clean-water infrastructure design under peat-soil constraints, whereas Parabi et al. [15] delineated peatland extent and land-use change through geospatial demarcation. Neither prior study quantified below-ground carbon stocks, performed SEM-EDX microstructural characterisation, or produced carbon-density maps, which constitute the dataset and analytical contribution reported here. To address this gap, the present study pursues three objectives: (1) to quantify below-ground carbon stocks (MgC·ha⁻¹) in both districts through field-based peat core sampling and laboratory carbon analysis; (2) to characterise the microstructure and elemental composition of peat specimens using SEM-EDX analysis; and (3) to generate GIS-based spatial carbon stock distribution maps as a decision-support tool for urban spatial planning. To the best of our knowledge, the study thereby delivers the first district-level, multi-analytical carbon stock baseline for urban peatlands in Pontianak City, provides SEM-EDX-based confirmation of material identity at the micro-scale for a representative specimen, and produces GIS carbon density maps directly applicable to urban spatial planning in Kalimantan Barat Province.

3.1 Field investigation and soil sampling results

Field verification across four candidate districts confirmed a continuous peat layer traced by hand auger to the organic–mineral boundary—in North Pontianak (1,732.16 ha) and Southeast Pontianak (579.80 ha) only; no comparable peat layer was identified in Pontianak City or East Pontianak, which were accordingly excluded from further sampling. Sampling density was scaled to each district’s confirmed peatland area, yielding 37 georeferenced stations (23 North, 14 Southeast) for peat depth profiling, in situ pH, Von Post humification assessment, and laboratory determination of organic content. Table 1 reports these organic content results by district: North Pontianak, consistent with its intact-to-semi-degraded peat setting (Section 2.1), returns a mean of 87.59% with approximately 96% of points meeting the ASTM D4427 ≥75% threshold, whereas Southeast Pontianak—already characterised in Section 2.1 as transitional development over disturbed residual peat swamps—returns a lower and more variable mean of 63.60%, with approximately 67% of points meeting the threshold. This organic-content pattern is consistent with, rather than a departure from, the disturbed peat setting Southeast Pontianak was selected to represent.

Table 1. Organic content and peat classification results across candidate districts

3.2 Peat depth investigation results

Peat depth was profiled at each station using a hand auger (10 m capacity) to the organic–mineral boundary, sampling each maturity horizon; results appear in Table 2, with the resulting extent compared against preliminary IMDGNP estimates in Table 3.

North Pontianak depths ranged from 2.00–4.30 m (Table 2; one individual auger point recorded 1.05 m and is treated as a local outlier rather than the district range boundary), dominated by the Deep class (1,597.42 ha) per Government Regulation No. 57/2016 (>3.00 m = ecosystem protection; ≤3.00 m = agricultural productivity). Southeast Pontianak depths ranged from 1.25–3.25 m, with no Very Deep deposits, a shallower profile reflecting proximity to urban drainage and consequent peat subsidence through oxidative decomposition.

Table 2. Peat layer depth range and ecosystem function by district

Table 3. Areal distribution of peat depth classes by district

3.3 Von Post humification assessment and peat maturity

Von Post humification results for all stations, assessed at successive depth intervals, are summarised by each district's predominant class in Table 4, with station-level variation detailed below.

All stations in both districts predominantly yielded H4–H5 classification (hemic peat, fibre content 33–67%, ASTM D1997), denoting slightly to moderately decomposed peat. Of 23 North Pontianak stations, 20 are hemic, 2 sapric, and 1 fibric; of 14 Southeast Pontianak stations, 11 are hemic, 2 fibric, and 1 sapric. Hemic peat accounts for 98.5% of the combined 2,311.96 ha extent, making it the primary driver of below-ground carbon stocks. Thiessen (Voronoi) zonation of these per-point maturity classifications, following the same procedure described in Section 2.4, yielded the maturity-class areal extents in Table 5.

Table 4. Von Post scale of humification examination results (predominant classification by district)

Table 5. Areal extent of peat maturity (fibre content) classes by district

3.4 Scanning Electron Microscopy coupled with Energy-Dispersive X-Ray Spectroscopy morphological analysis of peat soil

SEM micrographs at 500× and 1000× (Figure 6) reveal a dual-porosity architecture in North Pontianak peat—macropores between fibre elements and micropores within the fibre matrix—consistent with hemic classification. This interpretation is limited strictly to pore morphology and elemental composition: the macropore/micropore differentiation reflects the fibrous, partially decomposed structure characteristic of hemic peat, consistent with the Von Post humification class assigned in the field. No dissolved organic carbon (DOC) measurements were made in this study, so no inference is drawn from pore structure to DOC concentration or hydraulic transport behaviour; the bulk organic carbon density used in the carbon stock calculation is derived independently from LOI measurements (Section 2.2), not from this SEM pore-structure observation. EDX elemental composition results are presented in Table 6.

6.png

Figure 6. Scanning Electron Microscopy (SEM) micrographs of North Pontianak peat soil at (a) 500× and (b) 1000× magnification

Table 6. Elemental composition of North Pontianak District peat soil by Scanning Electron Microscopy coupled with Energy-Dispersive X-Ray Spectroscopy (SEM-EDX) analysis

3.5 Carbon stock based on peat depth classification

Below-ground carbon stocks were stratified by peat depth class (Moderate, Deep, Very Deep) per Government Regulation No. 57/2016, using the bulk density, organic carbon fraction, and layer thickness values described in Research Methods. Table 7 presents the resulting carbon stock and density by depth classification, and Figure 7 maps this depth-based carbon stock classification spatially across both districts.

Table 7. Carbon stock and carbon density by peat depth classification

7.png

Figure 7. Spatial mapping of carbon stock based on peat depth classification

Deep peat (200–400 cm) is the principal carbon reservoir in both districts, with carbon densities (approximately 1,727 t C·ha⁻¹ Southeast; 1,423 t C·ha⁻¹ North) substantially exceeding the moderate-depth class; very deep deposits in North Pontianak (126.62 ha) reach approximately 1,669 t C·ha⁻¹. Combined totals of 850,989 t C (Southeast) and 2,490,875 t C (North) yield 3,341,864 t C at a mean density of approximately 1,446 t C·ha⁻¹ over 23.12 km²—near the 1,552.53 t C·ha⁻¹ reported by Septian et al. [32] for Jambi and well above degraded-soil values—confirming that Pontianak’s urban peatlands hold climatically significant carbon stocks; translating this stock size into an actual emissions risk, and any consequent groundwater or land-use management priority, would require further data on drainage depth, oxidation rate, and land-use change trajectory not measured in this study.

3.6 Carbon stock based on peat maturity classification

Maturity-based totals are 7.84% lower than depth-based estimates in Southeast Pontianak, 2.32% higher in North Pontianak, and 0.27% lower combined (Table 8)—a bounded discrepancy we report as an internal cross-check between the two classification schemes, not as confirmation of agreement between them. Both schemes classify the same 37 Thiessen zones over the same total area in each district (579.80 ha Southeast, 1,732.16 ha North) and draw on the same underlying point-level bulk density, organic carbon content, and depth measurements; the depth-based scheme groups zones by physical depth threshold per Government Regulation No. 57/2016, while the maturity-based scheme groups the same zones by Von Post humification class, so a given zone can be grouped with different stratum-mates under each scheme (for example, a zone classified Deep by depth may sit alongside different neighbouring zones than when it is classified Hemic by maturity). Because each stratum’s carbon stock is calculated from its own area-weighted mean bulk density, organic carbon fraction, and depth (Section 2.4) rather than from each zone’s exact individual value, a change of stratification axis changes which zones are averaged together and therefore changes the resulting stratum-level and district totals, even though both schemes classify an identical total area from identical point data. The larger discrepancy in Southeast Pontianak (7.84%) relative to North Pontianak (2.32%) is consistent with Southeast Pontianak’s smaller sample (14 versus 23 points) and its lower, more variable organic content (Table 1; Section 3.1), both of which make its stratum-average properties more sensitive to which zones fall into which group under each classification scheme. Table 9 breaks down carbon stock by maturity class: hemic peat dominates maturity-based carbon storage in both districts, contributing approximately 96% of Southeast Pontianak’s 784,279 t C and 99% of North Pontianak’s 2,548,645 t C, with fibric and sapric classes each under 4%. Figure 8 maps this maturity-based carbon-density classification spatially across both districts.

8.png

Figure 8. Spatial mapping of carbon stock based on peat maturity classification

Table 8. Reconciliation of depth-based and maturity-based carbon stock totals

Table 9. Below-ground carbon stock by peat maturity class

3.7 Spatial distribution of organic carbon

Figure 9 maps the spatial distribution of organic carbon content (%C) across both districts. Mean %C of approximately 53% in both Southeast and North Pontianak—statistically indistinguishable between districts at this precision—after Van Bemmelen conversion matches published Indonesian benchmarks (49–57%), confirming that district-level stock differences are driven primarily by deposit depth and spatial extent rather than carbon concentration.

9.png

Figure 9. Spatial distribution of organic carbon content

3.8 Implications for spatial planning and study limitations

Beyond the carbon stock totals reported in Sections 3.5–3.7, three further aspects govern how these results should be interpreted and applied: material identity, planning implications, and study limitations. SEM-EDX analysis of a single representative North Pontianak specimen confirmed the material identity of the tested peat: a dual-porosity fibre network and a carbon-dominated elemental composition (64.75 mass% C, 33.66 mass% O) with only trace Al, Si, and Ca, consistent with the organic-rich, mineralogically clean matrix expected of genuine peat under ASTM D4427. This corroborates the classification of the sampled material rather than validating the magnitude of the district-level carbon stock estimate, and was performed on a single representative North Pontianak specimen only (selected based on visually homogeneous peat material across sampling points); equivalent characterization for Southeast Pontianak was not performed (see limitations below). Carbon densities only marginally below pristine benchmarks indicate that Pontianak’s urban peatlands hold climatically significant carbon stocks despite ongoing development. This stock size is a necessary baseline for, but does not by itself quantify, the district’s emissions risk: translating it into an actual CO₂ emissions estimate would require further data on drainage depth, oxidation rate, and land-use change trajectory, none of which were measured in this study. With that caveat, the resulting spatial carbon density maps offer a baseline decision-support input for Pontianak City’s spatial planning, to be paired with such emissions-rate data before they inform specific groundwater or land-use management decisions.

Read together, the peat-depth classification map (Figure 7) and the carbon-density classification (Section 2.4) translate directly into a planning-oriented decision framework. Under Government Regulation No. 57/2016, areas mapped as Deep or Very Deep (>3.00 m; Table 3) already carry an ecosystem-protection function designation, so Figure 7 can be read as a protection-versus-agricultural-use zoning overlay for the study districts. Overlaying this with the carbon-density classification (Section 2.4) identifies a composite priority-conservation zone—areas that are both Very Deep and High-density (126.62 ha in North Pontianak)—where the regulatory protection mandate and the carbon-loss consequence of disturbance reinforce one another; we present this composite zone as an illustrative application of the mapped data to planning practice, not as a zoning determination, which remains the statutory responsibility of Pontianak City’s spatial planning authority. For groundwater and drainage-related decisions specifically, the same depth and density maps indicate where drainage control would carry the largest carbon-stock consequence, offering a spatial prioritisation for the existing piezometric network (Section 2.4) rather than a substitute for it. For foundation and construction planning, the general engineering behaviour of deep, compressible tropical peat—typically requiring piled or ground-improved foundations and staged, controlled dewatering to limit oxidative subsidence, consistent with prior peat-soil engineering design experience in this delta [14]—applies most directly to the Deep and Very Deep zones mapped here; we present this as a qualitative planning flag rather than a site-specific geotechnical recommendation, since compressibility, bearing capacity, and settlement parameters were outside the scope of this study and require dedicated geotechnical investigation before any design decision. Finally, the zonation structure itself indicates where future monitoring would most reduce uncertainty: because each Thiessen zone’s value is assigned from a single point, uncertainty is concentrated at zone boundaries and in the more sparsely sampled Southeast Pontianak District (14 points over 579.80 ha, versus 23 points over 1,732.16 ha in North Pontianak; Section 2.1); within a constrained monitoring budget, additional points sited near existing zone boundaries or within Southeast Pontianak’s larger zones would reduce district-level uncertainty more than equivalent points added inside already well-constrained, homogeneous zones.

This study has four limitations. First, although bulk density and organic content were measured in triplicate at each point and depth interval, replicate-level measurement uncertainty was not propagated into a formal error envelope; because carbon stock mapping uses deterministic Thiessen (Voronoi) zonation rather than a fitted geostatistical interpolator, there is no variogram, nugget, sill, range, or cross-validation statistic to report—any reader reconstructs an identical zone map from the same point data. The remaining, measurement-level uncertainty instead propagates through the carbon stock formula (Eq. (2)) as (σC_stock/C_stock)² ≈ (σρ_b/ρ_b)² + (σ%C_organic/%C_organic)² + (σD/D)², where σρ_b, σ%C_organic, and σD are the standard deviations of the triplicate bulk density, LOI-derived organic carbon, and depth measurements at each point; these per-point standard deviations were recorded during triplicate testing but were not compiled or reported at the district level in this manuscript. We flag this explicitly as the outstanding step required to convert the point-estimate carbon stocks reported here into a quantified uncertainty envelope, rather than leaving uncertainty an unquantified caveat. Second, the field protocol did not define an explicit, documented rule for confirming that every sampling point was clear of built-up structures beyond in-field relocation to the nearest accessible ground (Section 2.1); no station was knowingly sited inside a built structure, but this was not formally logged point-by-point. Third, SEM-EDX characterization—used here only to confirm that the tested material is genuine, mineralogically clean peat rather than to validate the carbon stock magnitude—was performed on a single representative North Pontianak specimen (n = 1), selected for visual homogeneity with the wider sampled material rather than as a statistically replicated series. North Pontianak is accordingly treated as the primary candidate district for carbon stock calculation in this study, with Southeast Pontianak retained as a comparison case rather than an equally validated calculation basis; the LOI-based organic-carbon conversion therefore rests on a stronger evidentiary footing for North Pontianak than for Southeast Pontianak, where Table 1 shows a lower mean organic content. The 579.80 ha and 1,732.16 ha calculation areas are not themselves defined by this organic-content threshold: both were field-confirmed as continuous peat by auger profiling to the organic–mineral boundary (Section 3.1), independent of the point-level organic-content result subsequently measured in the laboratory. Organic content therefore functions in this study as a measured input to the carbon-stock formula (Eq. (2)) at each point, not as an area-inclusion criterion; Southeast Pontianak’s lower and more variable organic content is consequently carried through the reported stock as the elevated, district-specific uncertainty registered above, rather than as an exclusion of any part of its confirmed peatland extent. Fourth, the districts were sampled at different densities (23 stations/1,732.16 ha North versus 14 stations/579.80 ha Southeast) proportional to their confirmed peatland extent rather than a fixed quota per district, so the shallower Southeast profile cannot be fully disentangled from this coverage difference.