Soil and Climate Controls on the Economic Value of Forest Carbon in Northeast China

Abstract

Broad-scale assessments often track forest productivity, yet they rarely quantify how soil conditions determine whether these gains persist as long-lived carbon and generate measurable economic value. This study focused on Northeast China, where forests include boreal coniferous stands dominated by Dahurian larch, temperate conifer–broadleaf mixed forests with Korean pine, and temperate deciduous broadleaf forests dominated by Mongolian oak. We combined GLASS net primary productivity and ESA CCI Land Cover to delineate forest pixels, used 2000 to 2005 as the baseline, and converted productivity anomalies into pixel level carbon economic value using a consistent pricing rule. Forest NPP increased significantly during 2000 to 2018 (slope = 1.57, p = 0.019), and carbon economic value also increased over time during 2006 to 2018 (slope = 2.24, p = 0.002), with the highest values in core mountain forests and lower values in the western forest–grassland transition zone. Correlation analysis, explainable random forests, and variance partitioning characterized spatial and temporal dynamics from 2000 to 2018 and identified environmental controls. Carbon value increased over time and showed marked spatial heterogeneity that mirrored productivity patterns in core mountain forests. Climate was the dominant predictor of value, while higher soil pH and clay content were negatively associated with value. The random forest model explained about 70% of the variance in carbon value (R² = 0.695), and variance partitioning indicated substantial unique and joint contributions from climate and soil alongside secondary topographic effects. The automatable framework enables periodic updates with new satellite composites, supports ecological compensation zoning, and informs soil-oriented interventions that enhance the monetized value of forest carbon sinks in data-limited regions.

1. Introduction

With accelerating climate warming and converging carbon neutrality goals, terrestrial carbon sinks are central to stabilizing the global carbon cycle [1,2,3]. Forests, the largest terrestrial carbon pool, fix carbon via photosynthesis, and NPP constrains the carbon available for storage and turnover [4]. Northeast China (Greater and Lesser Khingan and the Changbai ranges) is a key node in the Northeast Asian carbon cycle with implications for national carbon accounting [5,6]. Yet most broad-scale assessments focus on productivity trends and rarely assess whether gains persist as long-lived carbon, leaving a key step in sink formation unresolved.

Long-term variations in forest sinks arise from the joint influence of pedoclimatic situations and human activities [7,8]. Warming, shifts in precipitation regimes, and extremes alter growing season length, the photosynthesis–respiration balance, and phenology, thereby regulating interannual and decadal NPP [9,10]. The conversion of climate-driven productivity signals into realized carbon accumulation is mediated by soils. Precipitation-controlled soil moisture and its seasonal timing govern transpiration–photosynthesis coupling and water-use efficiency, and soil physicochemical properties, including clay content, bulk density, pH, and the SOC background, regulate rhizosphere supply, pore structure, and heterotrophic respiration [11]. Where soils buffer hydroclimatic variability and sustain favorable rooting environments, productivity increments are more likely to persist as carbon gains. Where soils are short of moisture or have poor structure, the same climate shifts can be lost before they turn into stored carbon. This soil-controlled bottleneck is rarely accounted for in regional carbon valuation.

Human actions also reshape the link from productivity to storage. Logging, land conversion, and urban growth reduce forest carbon, while protection, cropland-to-forest programs, and silviculture improve stand structure and soils [12]. In such a managed landscape, the value drawn from carbon stocks should depend on how climate and soils pass productivity gains into stored carbon. This motivates three questions for Northeast China. First, to what extent do temperature, precipitation, and aridity sort the spatial pattern of carbon economic value? Second, do soil properties such as pH and clay content impose constraints that limit the portion of NPP gains that becomes long-lived carbon? Third, do climate and soils act independently or together with topography? Framing valuation around these controls is essential as China’s emissions trading system links stock changes to monetary value for ecological compensation and project choice [13,14].

Economic valuation of forest carbon sinks is commonly conducted by converting estimated carbon sequestration, or changes in carbon stocks, into monetary value using either carbon market prices or the social cost of carbon, sometimes combined with discounting to support net present value-based comparisons across scenarios. Such approaches have been applied across diverse regions and forest types, including protected areas and tropical dry forests, and are often operationalized through spatial ecosystem service models such as InVEST, that value sequestration or loss under alternative scenarios [15]. For example, Pache et al. [16] valued carbon storage and sequestration in Retezat National Park, Romania, by applying the InVEST Carbon model to 30 m land using maps and forest management plan inventories, and reported a net present value of USD 1,706,070.28 for carbon sequestration over 2019 to 2029 under a voluntary market price of USD 3 per tCO2e, while showing that values increase substantially under higher carbon price benchmarks; their modeled estimates were further evaluated using terrestrial laser scanning plots. However, these studies typically emphasize carbon quantity and pricing assumptions, while the extent to which soil conditions regulate the conversion of productivity gains into more persistent carbon pools remains less resolved. This matters because soil mineral association, texture, and pH can strongly shape carbon persistence and turnover, affecting whether gains in productivity are retained rather than rapidly returned to the atmosphere [17]. Against this backdrop, the present study advances a soil-aware pixel level valuation framework for Northeast China by linking productivity anomalies to relative carbon economic value under a consistent pricing rule and quantifying how climate, soils, and topography jointly control the resulting spatial pattern of value.

A central gap is that broad-scale assessments track productivity but rarely specify the soil and climate conditions under which gains persist as long-lived carbon and acquire economic value. Focusing on Northeastern China from 2000 to 2018, this study maps the spatiotemporal pattern of forest productivity, quantifies the associated change in relative carbon economic value (ΔEi), and evaluates how climate, topography, and soils together organize the spatial distribution of ΔEi. It further tests whether core mountain forests are more stable than ecotonal margins, whether wetter climates support higher value for a given productivity gain, and whether soil properties such as pH, clay content, and background SOC limit the share of productivity gains that becomes durable value. By embedding valuation within this soil-aware framing, the work provides actionable evidence for carbon-sink management, ecological compensation zoning, and informed participation in regional carbon markets.

2. Study Site

The study area comprises forestlands in Northeast China, covering Heilongjiang, Jilin, and Liaoning Provinces in full and the northeastern part of the Inner Mongolia Autonomous Region (including Hulunbuir and Hinggan League) (Figure 1). It extends from 38°42′ N to 53°31′ N and from 115°52′ E to 135°09′ E. This region contains China’s largest temperate conifer–broadleaf mixed forests and cool-temperate coniferous forests and is a core carbon-sink zone in Northeast Asia [18].

Prevailing soils vary strongly with topography and hydroclimate across Northeast China. In the major mountain forest belts, including the Greater Khingan, Lesser Khingan, and Changbai ranges, zonal forest soils are widespread, notably brown coniferous forest soils and dark brown forest soils, while podzolic soils and peat or bog soils occur locally in colder and wetter environments [19]. In lower elevation plains and at forest margins, soils shift toward chernozems and other black soils, and meadow soils are common in areas with impeded drainage [19]. This soil mosaic provides a basis for the observed spatial contrasts in soil pH, texture, and SOC that may constrain the conversion of productivity gains into durable carbon and associated economic value.

Topography slopes from west to east and is higher in the north and gentler toward the south. The climate spans a temperate continental monsoon regime transitioning to a cool-temperate zone, with pronounced spatiotemporal variability in water and energy. Mean annual temperature ranges from −5 to 6 °C. Growing degree days above 10 °C total 2000–3500 °C·day. The frost-free period is 100–160 days, and permafrost occurs in the northern sector. Mean annual precipitation is 300–1000 mm, higher in the southeast than the northwest, with 60–70 percent falling during June to August. From 2000 to 2018, mean annual temperature increased at 0.23 °C per decade, precipitation showed strong interannual variability with coefficients of variation of 12–18 percent, and the region exhibited a warming and drying tendency [18].

Forest is the dominant land use at 62 percent, followed by cropland at 25 percent, grassland at 8 percent, and built-up land at 3 percent. Vegetation types are well differentiated: the northern Greater Khingan hosts cool-temperate coniferous forests dominated by Dahurian larch; the Lesser Khingan and Changbai Mountains support temperate conifer–broadleaf mixed forests with Korean pine, linden, and associated species; and the southern sector features temperate deciduous broadleaf forests dominated by Mongolian oak. Since 2000, ecological restoration programs, including the Natural Forest Protection Program and the Conversion of Cropland to Forest Program, have increased forest area by 0.32 percent per year, raising forest cover from 48.5 to 51.2 percent and providing a stable basis for carbon-sink function [12].

2.1. Data Sources

Net primary productivity (NPP) data were obtained from the Global Land Surface Satellite (GLASS) evapotranspiration and vegetation productivity product, covering 2000–2018. GLASS integrates multiple satellite observations such as MODIS and AVHRR and derives NPP using an improved light-use-efficiency model [20]. The spatial resolution is 500 m and units are g C m⁻² yr⁻¹. The dataset provides strong temporal continuity, high spatial consistency, and good compatibility with ecosystem processes, and has been widely used for global and regional vegetation carbon cycle studies.

To extract NPP for the forested area of Northeast China, we applied a two-step masking procedure to the raw GLASS NPP:

Administrative masking. Using vector boundaries for Northeast China (Heilongjiang, Jilin, and Liaoning Provinces, and northeastern Inner Mongolia), we clipped GLASS rasters in ArcGIS (version 10.2) to remove pixels outside the study region and ensure spatial alignment with administrative limits.

Land-cover masking. Using the ESA Climate Change Initiative Land Cover product (ESA CCI-LC, 300 m, annual), we selected forest classes from the level-2 legend, including evergreen needleleaf, deciduous needleleaf, evergreen broadleaf, deciduous broadleaf, and mixed forests [21]. This forest mask was applied to the regional GLASS NPP to obtain an annual time series of forest-only NPP for 2000–2018, excluding cropland, grassland, and built-up areas.

The ESA CCI-LC product combines multiple satellite sensors, including Envisat MERIS and Sentinel-3 OLCI, and is produced with radiometric correction, geometric refinement, and optimized classification algorithms [22]. The overall classification accuracy exceeds 85 percent, supporting precise delineation of forest extent in the study area and providing a reliable basis for the forest-specific extraction of NPP.

Soil properties were derived from the SoilGrids 250 m product (SoilGrids v2.0) provided by ISRIC World Soil Information (https://maps.isric.org/ and https://soilgrids.org/, accessed on 20 January 2025). SoilGrids provides global predictions of key soil attributes at 250 m resolution across six standard depth intervals. We extracted clay content, soil organic carbon, soil pH, and bulk density layers and harmonized them to the spatial grid used for the NPP analysis through reprojection and resampling, followed by masking to forest pixels.

2.2. Forest Carbon Dynamic

2.2.1. Conversion from ΔNPP to Carbon-Stock Difference (ΔC)

Based on annual ΔNPP for 2006–2018 (the difference between yearly NPP and the 2000–2005 mean), we converted from productivity units to the carbon-stock unit commonly used in carbon-sink studies as follows:

ΔC = ΔNPP × k

(1)

where ΔC is the grid-cell carbon-stock difference relative to the 2000–2005 mean, in t C ha⁻¹ yr⁻¹. Positive values indicate a carbon-sink increase and negative values indicate a decrease. ΔNPP is the difference between annual NPP and the 2000–2005 mean, in g C m⁻² yr⁻¹. $k$ is a unit-conversion factor equal to 0.01 (dimensionless). The derivation is as follows: 1 ha = 10⁴ m² and 1 t = 10⁶ g, therefore 1 g C m⁻² yr⁻¹ = 1 g C/m²/yr = (1 × 10⁻⁶ t C)/(1 × 10⁻⁴ ha) = 0.01 t C/ha/yr. This factor harmonizes units to allow direct conversion from productivity to carbon stock.

2.2.2. Pixel Level Carbon Economic Value (Conversion from ΔC to $E_{\mathit{i}}$)

Using ΔC and a carbon price parameter, we quantified the monetary contribution of the carbon sink for each pixel:

Ei = ΔCi × P

(2)

where $E_i$ is the carbon economic value of pixel $i$ relative to the 2000–2005 mean, in CNY ha⁻¹ yr⁻¹. $\mathsf{\Delta }{C}_i$ is the carbon-stock difference for pixel $i$ from Section 2.2.1, in t C ha⁻¹ yr⁻¹. $P$ is the carbon price. We adopt the national carbon emission allowance (CEA) mid-price for September 2025, 70.42 CNY per t C, derived from our center’s modeling, with an expected buy price of 68.46 CNY per t C and sell price of 72.38 CNY per t C, which offers market representativeness and stability. For a cost-side comparison, Torres, Marchant, Lovett, Smart, and Tipper [14] reported sequestration costs for agroforestry afforestation and reforestation in Asia of USD 8–32 per t C (about CNY 56–224 per t C at 2010 exchange rates), highlighting the strong effects of rotation length and labor costs; these can be used for sensitivity analysis. Interpretation of signs: $E_i>0$ indicates economic gains relative to the baseline, and $E_i<0$ indicates economic losses.

2.2.3. Identifying Drivers of Carbon Economic Value

We used correlation analysis to identify environmental drivers of the spatial pattern of forest carbon economic value. Climate variables included mean annual temperature, mean annual precipitation, and aridity index. Topographic variables included elevation, slope, and aspect. Soil variables included clay content, pH, SOC, and bulk density. A random forest model was applied to assess the relative importance of these factors. Finally, variance partitioning was used to quantify the independent and joint contributions of climate, soil, and topography. All analyses were conducted in R.

3. Results

3.1. Spatiotemporal Patterns of NPP in Northeast China

Forest NPP in Northeast China exhibits pronounced regional contrasts (Figure 2). Overall, higher NPP values occur in the northeast and southeast, shown as yellow to orange tones, corresponding to mountainous forest regions such as the Changbai Mountains and the Lesser Khingan, where growing conditions are more favorable, stand structures are more complex, and productivity is higher. In contrast, the west and parts of the central region display lower NPP, shown in green, largely corresponding to forest–grassland ecotones in the western Greater Khingan and to stands of lower quality where precipitation and heat limit productivity.

From 2000 to 2018, forest NPP shows clear interannual fluctuations (Figure 2). For example, in 2003 and 2004, high-value areas (yellow and orange) expanded, indicating climate conditions favorable for growth and an overall increase in NPP. In some years, such as 2005 and 2009, high-value areas contracted and green areas expanded, reflecting reduced NPP that may be associated with decreased precipitation, temperature anomalies, or disturbances such as pests or management activities.

Comparing individual years with the 2000–2018 mean shows that high-value areas remained relatively stable in the core northeastern and southeastern forest regions (Figure 2). Although interannual variability is evident, there is no persistent monotonic increase or decrease at this descriptive stage, suggesting overall stability during the study period, while the year-to-year swings highlight sensitivity to external drivers.

From the time-series trend of forest NPP during 2000–2018, NPP shows a significant upward trajectory (slope = 1.57, p = 0.019 < 0.05), indicating statistical significance (Figure 3). Superimposed on this increase are marked interannual oscillations, including brief declines around 2003–2004 and 2008. These fluctuations likely reflect the combined influences of climatic anomalies, such as reduced precipitation and extreme temperatures, and of human interventions, such as local harvesting and tending. The relatively narrow 95% confidence interval (green shading) indicates that the fitted linear trend is robust rather than driven by random noise. Overall, forest ecosystem productivity and associated carbon-sink potential in Northeast China increased appreciably over 2000–2018.

3.2. Spatial Dynamics of Forest ΔNPP in Northeast China (2000–2018)

Using the 2000–2005 mean NPP as a baseline (Figure 4a), the 2006–2018 anomalies (ΔNPP) display complex and regionally differentiated spatial patterns. Across years (panels b–o), ΔNPP is far from uniform: some areas show positive anomalies (warm colors) indicating increases over the baseline and enhanced sink potential, while others show negative anomalies (cool colors) indicating decreases.

In 2006 (Figure 4b), positive ΔNPP dominates the northeast and parts of the central region, likely reflecting favorable hydrothermal conditions that year, while parts of the west show negative ΔNPP, possibly linked to precipitation deficits. In 2008 (Figure 4d), large portions of the central region exhibit positive ΔNPP, whereas some northeastern patches turn negative. In 2012 (Figure 4h), positive ΔNPP concentrates in the south, with negative anomalies in parts of the north, highlighting a north–south contrast. By 2018 (Figure 4n), positive ΔNPP prevails in the northeast and southeast, indicating strengthened sink potential in core forest regions, while portions of the west remain negative relative to the baseline.

Taken together with the baseline map, core high-NPP regions such as Changbai and the Lesser Khingan show positive ΔNPP in most years, implying persistently higher productivity and more stable sink behavior, whereas edge and ecotonal forests experience frequent sign changes and lower stability, being more strongly affected by climate and human disturbance. In summary, relative to 2000–2005, ΔNPP during 2006–2018 shows strong spatial heterogeneity, with core forests consistently above baseline and ecotones more volatile due to intrinsic site conditions and external drivers.

3.3. Spatial Dynamics of Forest Carbon Economic Value (ΔEi) in Northeast China (2000–2018)

Relative to the 2000–2005 baseline NPP (panel a), the 2006–2018 changes in carbon economic value (ΔEi) exhibit spatial patterns that closely track NPP dynamics and are strongly heterogeneous across the region (Figure 5). Areas with positive ΔEi (warm colors) deliver higher carbon value than the baseline, indicating the enhanced economic contribution of the sink, while negative ΔEi (cool colors) indicates reduced value.

In 2006 (Figure 5b), positive ΔEi dominates the northeast and parts of the central region, consistent with higher-than-baseline NPP and increased carbon stock; the western sector tends to be negative. In 2008 (Figure 5d), a broad central belt shows positive ΔEi, reflecting productivity-driven gains, while some northeastern patches are negative, paralleling local NPP declines. By 2018 (Figure 5n), ΔEi is mainly positive in the northeast and southeast, indicating the sustained enhancement of economic contribution in core forests where NPP has remained above baseline, while parts of the west remain negative.

Overall, core high-NPP regions such as Changbai and the Lesser Khingan register positive ΔEi in most years, suggesting a stable above-baseline value, whereas edge and ecotonal forests show frequent sign changes due to combined effects of NPP variability and price factors. In sum, relative to 2000–2005, ΔEi is strongly associated with NPP and yet is spatially differentiated, with persistent gains in core forests and higher volatility in ecotones under the joint influence of biophysical and economic drivers.

From the time-series plot of ΔEi for 2006–2018, ΔEi exhibits clear and statistically significant temporal behavior (Figure 6). There is evident interannual variability, as shown by the boxplots for each year and the changing standard deviations. The linear fit (“Linear Fit of NPP” in the figure) indicates a significant upward trend in ΔEi (slope = 2.24, p = 0.002). Although ΔEi declines temporarily or even turns negative around 2008–2010, the long-term trajectory is upward, indicating a marked improvement in carbon economic value relative to the baseline. The narrow 95% confidence interval (pink shading) further supports the reliability of the trend and suggests that the economic contribution of the forest sink has strengthened over time, consistent with the long-term behavior of NPP and price-related factors.

3.4. Environmental Drivers of the Spatial Dynamics of Forest Carbon Economic Value

Correlation analysis shows that ΔEi rises with mean annual temperature, mean annual precipitation, and the aridity index, highlighting the dominant role of regional climate water–energy balance in structuring the spatial pattern of carbon economic value across Northeast China forests (Figure 7a). In contrast, higher soil pH and clay content are associated with lower ΔEi, implying that more alkaline and finer-textured soils limit the conversion of productivity gains into monetizable carbon benefits, potentially through constraints on root inputs and stabilization pathways. The random forest explains about 70% of the variance in ΔEi (R² = 0.695) and consistently ranks climate variables as the most informative predictors, led by temperature, precipitation, and aridity, followed by topographic metrics such as elevation and slope, while soil variables contribute additional but smaller predictive power (Figure 7a).

Variance partitioning further resolves these effects by attributing unique fractions of 9% to climate, 15% to soil, and 1% to geographic position, along with sizable shared fractions for climate–soil (15%), climate–geographic (4%), soil–geographic (3%), and a three-way overlap of 3%, leaving 50% residuals that likely reflect local management, stand structure, disturbance history, and pricing uncertainties not captured by the current predictors (Figure 7b). Taken together, climate emerges as the primary control, soil provides a comparable unique share and interacts strongly with climate, and topography and geographic position play secondary roles while helping to capture spatial structure.

4. Discussion

4.1. Spatiotemporal Linkage Between NPP and Carbon Economic Value: Coupling of Ecological Processes and Market Valuation

As the core descriptor of ecosystem productivity, NPP captures the realized capacity of vegetation to fix carbon through photosynthesis across space and time [23,24]. Carbon economic value is the monetary mapping of this ecological process under market rules. Their association reveals how the ecological and economic subsystems are coupled. Interpreting pixel-scale carbon economic value derived from productivity anomalies requires recognizing the transformation chain from plant production to persistent carbon storage. NPP represents net plant carbon gain after autotrophic respiration, but the fraction that becomes long-lived carbon depends on allocation to woody biomass and soil pools and on carbon residence time. Net ecosystem production is defined as NPP minus heterotrophic respiration, and heterotrophic respiration can respond differently than NPP to climatic variation and management, which can lead to divergence between NPP and NEP across space and time [25]. Accordingly, NPP anomalies should be viewed as a proxy for potential additional carbon inputs rather than a direct estimate of net carbon sequestration, which motivates coupling satellite-derived productivity with process-based carbon cycle models such as CENTURY or related frameworks to represent carbon allocation, litter inputs, and soil carbon turnover more explicitly [26].

Spatially, high-NPP zones in Northeast China, such as the Changbai Mountains and the core of the Lesser Khingan, benefit from superior site conditions including soil fertility and water retention and from relatively stable stand structures dominated by species such as Korean pine and Dahurian larch. In most years from 2006 to 2018, these areas maintain positive ΔNPP. Under a relatively stable price environment in which this study adopts the CEA mid-price of CNY 70 per t C, increases in carbon stock are directly valued as ΔEi. This forms a transmission chain from productivity advantage to larger carbon increments to higher economic value. In contrast, edge forests and forest–grassland ecotones are more exposed to external disturbances [27]. Uneven interannual rainfall distribution that triggers water stress in the west and temperature anomalies that suppress spring leaf can directly depress photosynthesis and induce NPP volatility [28]. Local land use changes and management interventions also alter canopy cover and growth status, producing frequent sign switches in ΔNPP and amplifying spatial heterogeneity in ΔEi. This highlights a pathway from ecological fragility to unstable sink capacity to volatile economic value.

Temporally, NPP increases significantly from 2000 to 2018. This long-term rise reflects joint responses to global change, such as a longer growing season under warming, and to conservation programs such as natural forest protection and cropland-to-forest conversion [29]. These strengthen the material basis for gains in carbon economic value. Although extreme climate events and other factors produce year-to-year swings in NPP [30], ΔEi still rises gradually in the linear fit. This indicates a resilience-based conversion in which community succession and species compensation enable ecosystems to overcome short-term shocks and to convert sustained carbon fixation into economic returns over longer periods. The narrow 95 percent confidence interval further supports the statistical robustness of this ecological-to-economic linkage and shows that an NPP-based valuation framework is stable enough to support regional assessment, reinforcing the logic that long-term ecological trends translate into steadily increasing economic value.

4.2. Drivers of Carbon Economic Value in the Study Region: Coordination Between Natural Resilience and Human Regulation

The evolution of carbon economic value in Northeast China reflects the interaction between ecosystem resilience and proactive human regulation. Despite a warming and drying tendency that poses potential stress [31,32], NPP increases during the study period, indicating strong resilience. At the species level, dominant trees such as Dahurian larch and Korean pine are adapted to cool-temperate climates. The lengthening of the growing season can partially offset reduced precipitation [33]. At the community level, mixed conifer–broadleaf and cool-temperate coniferous forests have high diversity and complex structure, which enhances resource capture and disturbance resistance [34]. When hydrothermal conditions align with plant growth, for example, abundant summer rain and moderate temperature, NPP rises markedly [29] and ΔEi increases. Conversely, hot and dry summers depress NPP and reduce economic value. These pulse-type climatic effects are a primary natural cause of interannual fluctuations in ΔEi.

Human factors exert a leading regulatory role relative to the background influence of climate. Since 2000, large-scale restoration programs have reshaped the ecological baseline of Northeast forests in terms of both quantity and quality [34]. Restrictions on logging and conversion have expanded forest area, providing more substrate for carbon fixation. Tending and rehabilitation of degraded stands have improved structural integrity and forest health, allowing NPP to remain high and to increase over time, which underpins growth in economic value. Cost studies for afforestation and reforestation in agroforestry systems in Asia provide a complementary perspective on valuation and confirm the central role of human policies that guide ecological investment and market mechanisms that price carbon services [14]. This establishes a complete chain from regulation to enhanced ecosystem services to market realization.

4.3. Limitations and Future Research Directions

This study provides a spatially explicit assessment of forest carbon economic value and its environmental controls in Northeast China, but several limitations should be acknowledged. First, the valuation framework uses productivity anomalies combined with a fixed carbon price, which simplifies the full ecological-to-economic conversion pathway. In particular, net ecosystem production depends not only on plant carbon gains but also on carbon losses through heterotrophic respiration and decomposition, so approaches based on NPP alone can overestimate net carbon uptake when respiratory losses increase [25]. In addition, carbon price assumptions affect the magnitude and temporal dynamics of estimated value, yet carbon markets can exhibit substantial price volatility through time [35]. Future work will therefore couple remotely sensed productivity with process-based carbon cycle models such as CENTURY or DayCent to represent carbon allocation, litter inputs, and soil carbon turnover, enabling more robust estimates of NEP and more defensible valuation [26]. Dynamic pricing schemes, including time varying carbon market prices and scenario-based price trajectories, will also be incorporated to quantify the sensitivity of carbon economic value to price uncertainty [35].

Second, driver attribution in the current analysis is primarily statistical and does not explicitly resolve nonlinear threshold behavior or the causal effects of human interventions. Soil and hydroclimate controls can be non-additive and may show thresholds, which motivates combining explainable machine learning with targeted field or controlled experiments to identify critical ranges of soil pH and hydrothermal conditions. For program evaluation, quasi-experimental designs such as difference in differences can help separate heterogeneous impacts across ecological engineering projects while accounting for background trends and serial correlation [36]. In parallel, human pressures will be quantified more directly by integrating proxies such as nighttime lights, together with additional disturbance indicators where available, to represent urban expansion and other anthropogenic disturbances that may erode carbon-sink persistence [37].

Third, data limitations constrain both accuracy and timeliness. Remotely sensed NPP products can be less reliable in complex terrain, and global soil products such as SoilGrids, although valuable for regional analyses, may not capture fine-scale edaphic heterogeneity and carry quantified prediction uncertainty. To reduce these uncertainties, future work will improve NPP estimation through the fusion of Landsat and MODIS or related spatiotemporal fusion approaches and will strengthen links between satellite pixels and plot measurements using representative forest stands [38]. The analysis period will also be extended to 2025 to better align with the policy context of China’s carbon peaking and carbon neutrality targets, improving the policy relevance of observed trends [39]. Finally, comparative applications in other major forest regions of China, such as southwest and southeast forest zones, will be conducted to test the generality of mechanisms and to support region-specific strategies for enhancing carbon-sink value.

5. Conclusions

Forest productivity and its carbon economic value in Northeast China show clear spatial contrasts and significant temporal changes. Net primary productivity exhibits strong spatial heterogeneity, with consistently higher values in core mountain forest regions such as the Changbai Mountains and the Lesser Khingan Range and lower values in the western forest–grassland transition zone. From 2000 to 2018, forest NPP increased significantly (slope = 1.57, p = 0.019), although interannual variability indicates sensitivity to hydroclimatic fluctuations. Using the 2000 to 2005 period as a baseline, pixel-scale carbon economic value derived from productivity anomalies increased significantly during 2006 to 2018 (slope = 2.24, p = 0.002) and remained spatially concentrated in core forest regions. Explainable random forest models captured 69.5 percent of the spatial variance in carbon economic value (R² = 0.695), indicating that climate variables dominate the spatial pattern while soil properties, particularly pH and clay content, impose additional constraints. Collectively, these results demonstrate that spatial variation in the economic value of forest carbon sinks is jointly shaped by productivity dynamics and edaphic controls that influence carbon retention.