The study’s central quantities are two of the most important currencies in ecosystem science. The first is net primary productivity, or NPP, which measures the amount of carbon that vegetation fixes through photosynthesis minus the carbon plants respire for their own maintenance — in effect, the gross income of an ecosystem’s carbon economy. The second is carbon use efficiency, or CUE, the ratio of the carbon actually incorporated into plant tissue to the total carbon fixed. A high CUE means an ecosystem is a disciplined saver, converting its photosynthetic earnings into wood, roots, and leaves rather than burning them off as respiration. Together, these two metrics determine whether a landscape acts as a growing carbon sink or a leaky one, and both are exquisitely sensitive to the physical conditions of the root zone.

To track these variables across a region the size of a small continent, the researchers turned to the Community Land Model version 5.0, a state-of-the-art land surface model developed for Earth system simulation. CLM5.0 integrates soil thermal and hydrological processes, vegetation dynamics, and biogeochemical cycling, allowing the team to simulate NPP and CUE across the plateau from 1979 to 2020. The model was constrained and evaluated against an impressive array of observational data: long-term in situ soil temperature records archived at the National Tibetan Plateau Data Center, eddy-covariance measurements of net ecosystem productivity and heterotrophic respiration from the Haibei research station, satellite-derived MODIS net primary productivity products, and independent NPP estimates from the CASA and VPM remote sensing models. Soil property maps and plant functional type distributions provided the underlying substrate and vegetation structure.

The simulation revealed a striking hierarchy among the plateau’s ecosystems. Forests, concentrated in the warmer and wetter southeastern margins, were the most productive, with a median NPP of 963.52 grams of carbon per square meter per year. Shrublands, however, proved to be the most frugal, achieving the highest median carbon use efficiency at 0.53 — meaning more than half of the carbon they fixed was retained in biomass rather than lost to respiration. Grasslands, which blanket the vast majority of the plateau, showed the most dramatic long-term trajectories: their NPP climbed by 2.16 grams of carbon per square meter per year each year, while their CUE declined by 0.0009 per year. That combination — rising gross productivity but falling efficiency — hints at a subtle but consequential shift in how these alpine ecosystems are handling a warming climate.

The declining CUE of grasslands deserves particular attention. Carbon use efficiency tends to fall when respiration outpaces photosynthesis, and warming accelerates both plant and microbial metabolism. Previous work cited by the team, including multi-model analyses of the Qinghai-Tibet Plateau, has suggested that continued climate change will erode the carbon use efficiency of terrestrial ecosystems across the region. The new results add a mechanistic dimension to that picture: the decline is not simply a response to warmer air, but is tightly coupled to the state of the soil, where permafrost thaw, altered drainage, and shifting moisture regimes reshape the environment in which roots and microbes operate. An ecosystem that fixes more carbon but spends proportionally more of it on respiration may accumulate biomass more slowly than its rising NPP would suggest.

To disentangle the relative influence of soil temperature and soil moisture, the researchers employed a random forest machine learning approach, a method well suited to capturing nonlinear relationships and interactions among many predictors. The analysis produced a clear seasonal signature. In spring and autumn — the shoulder seasons when the plateau’s vegetation is waking from or settling into dormancy — soil temperature was the dominant driver of NPP variation. This makes physical sense: during these periods, plant growth is often limited by cold soils, and the timing of spring thaw regulates when roots can access water and nutrients and when photosynthesis can resume. Earlier and warmer soil thawing has been shown to regulate the onset of growth in tundra and alpine biomes, and the new findings confirm that this temperature control extends across the plateau’s diverse vegetation types.

Summer told a different story. Once soils had warmed sufficiently, temperature ceased to be the binding constraint, and soil moisture took over as the key driver of net primary productivity. The plateau’s summer monsoon delivers the bulk of annual precipitation, and interannual variability in that moisture supply strongly modulates how much carbon vegetation can fix during the peak growing season. This finding aligns with global evidence that soil moisture exerts a large influence on long-term terrestrial carbon uptake, and it carries a sobering implication: as warming intensifies evapotranspiration and potentially destabilizes monsoon patterns, summer drought stress could become an increasingly important brake on the plateau’s carbon sink, even as longer and warmer growing seasons push in the opposite direction.

The vegetation-type analysis added yet another layer of nuance. For carbon use efficiency, soil temperature contributed more in grasslands, whereas soil moisture was the primary driver in shrublands and forests. These differences likely reflect the distinct ecophysiological strategies of the plant types involved. Grasses, with shallow root systems and rapid tissue turnover, are highly responsive to the thermal environment of the upper soil layers. Shrubs and trees, with deeper roots and woody structure, depend more heavily on sustained water availability, making their carbon allocation decisions more sensitive to moisture fluctuations. For land managers and carbon accountants, this means that a single regional parameterization of soil controls would misrepresent at least some of the plateau’s ecosystems; the rules of the carbon economy change with the vegetation.

Why does all of this matter beyond the plateau itself? The Tibetan Plateau, often called the Third Pole, holds one of the largest reservoirs of permafrost soil carbon outside the Arctic. As permafrost degrades and glaciers retreat, the soil hydrothermal regime is being rewritten, with consequences that ripple through the global carbon cycle. The study’s findings suggest that the trajectory of the plateau’s carbon sink will hinge on the coupled evolution of soil temperature and moisture — not on warming alone. A warmer but wetter plateau could continue to expand its vegetation carbon uptake, as the team’s earlier work on soil warming and wetting indicated, whereas a warmer and drier one could see productivity gains offset by respiration losses and declining efficiency. Regional carbon sink management and climate adaptation strategies, the authors argue, need to account for this hydrothermal coupling explicitly.

The study also demonstrates the growing power of combining process-based Earth system models with machine learning attribution. By simulating four decades of coupled soil and vegetation dynamics with CLM5.0 and then using random forests to apportion variance among drivers, the researchers achieved something neither approach could accomplish alone: a physically consistent reconstruction of ecosystem carbon behavior, paired with a data-driven ranking of its controls across seasons and vegetation types. The work was supported by the National Natural Science Foundation of China, the National Key R&D Program, and the Chinese Academy of Sciences, with simulations conducted using the Earth System Numerical Simulation Facility. As permafrost thaw accelerates across the world’s cold regions, this kind of quantitative attribution — knowing precisely when warmth matters and when water matters, and for which plants — will be essential for forecasting whether high-altitude and high-latitude ecosystems remain allies in the fight against climate change or become its amplifiers.

Subject of Research: Effects of soil temperature and moisture on vegetation carbon sequestration under climate warming on the Tibetan Plateau

Article Title: Quantifying the effects of soil hydrothermal conditions on vegetation carbon sequestration capacity under climate warming on the Tibetan Plateau

Article References: Peng, Q., Jia, B., Wang, L., Wu, R., & Huang, Q. (2026). Quantifying the effects of soil hydrothermal conditions on vegetation carbon sequestration capacity under climate warming on the Tibetan Plateau. Plant and Soil. https://doi.org/10.1007/s11104-026-09026-0

Image Credits: AI Generated

DOI: 10.1007/s11104-026-09026-0

Keywords: Tibetan Plateau, net primary productivity, carbon use efficiency, soil temperature, soil moisture, permafrost degradation, climate warming, CLM5.0, random forest, grasslands, carbon sink, alpine ecosystems

The research team, led by Wenhao Zhang and Xiaoyan Li of Beijing Normal University, chose the Qinghai Lake Basin as their natural laboratory for good reason. Covering nearly 30,000 square kilometers in the northeastern part of the plateau, the basin is split almost evenly between two very different frozen-ground regimes. Permafrost dominates the upstream, high-altitude areas, accounting for about 41.5 percent of the basin, while seasonally frozen soil covers the remaining 58.5 percent in the midstream, downstream, and lakeside plains. This natural gradient allowed the researchers to compare, within a single watershed, how permanently frozen ground and ground that freezes and thaws each year shape groundwater recharge in distinct ways.

To untangle these processes, the team built a comprehensive ecohydrological monitoring network across the basin. Between May and October 2025, they collected 481 samples in total: 90 of groundwater drawn from wells used by local herders, 306 of soil from three depth intervals — 0–30, 30–60, and 60–90 centimeters — and 85 of precipitation, including both rainfall and snow. Every sample was analyzed for its hydrogen and oxygen stable isotope composition using isotope ratio infrared spectrometry, with analytical precisions of ±0.5 per mil for deuterium and ±0.1 per mil for oxygen-18. Because every source of water — rain, snowmelt, soil moisture at different depths — carries a slightly different isotopic fingerprint, these measurements act like natural tracers, allowing the researchers to work out after the fact which waters ended up in the aquifer and how they got there.

The team divided the study window into three freeze–thaw periods based on sustained air and soil temperature thresholds: a thawing period in May and June, when soils transition from frozen to unfrozen; a thawed period from July to September, when the ground stays entirely unfrozen; and a freezing period in October, when temperatures drop back below zero. They then fed the isotope data into MixSIAR, a Bayesian mixing model that estimates the relative contribution of each potential water source to the groundwater while explicitly accounting for uncertainty. The results were striking. Across the entire freeze–thaw cycle, soil water was the dominant source of shallow groundwater recharge, contributing between 57.0 and 76.3 percent. Rainfall contributed 13.8 to 26.1 percent, and snowmelt — despite its dramatic visual presence on the landscape — contributed only 7.9 to 22.0 percent.

That snowmelt plays such a modest direct role may come as a surprise, and the authors offer two explanations. First, during the thawing period, air temperatures remain low, so snow melts slowly rather than releasing a sudden pulse of water. Second, under the basin’s intense solar radiation and strong winds, a significant share of snow water is lost to runoff or sublimation before it can infiltrate. Snow, in other words, functions largely as temporary storage on the surface rather than as an immediate source of groundwater. Instead, it is the water already held in the soil — moisture that fell as rain or snow in previous seasons and was released as the ground thawed — that does most of the work of refilling the shallow aquifer.

The study also revealed how the physical state of the soil controls the pathways water takes downward. During the thawing period, the active layer thaws only from the top, leaving a near-impermeable frozen layer beneath that blocks vertical infiltration. Meltwater and rain therefore pool in a transient saturated zone near the surface and migrate laterally as interflow, delaying groundwater recharge and often reaching the aquifer only indirectly, through riparian seepage and soil-layer redistribution. Once the thawed period arrives and this barrier disappears, the vertical connectivity of the soil profile increases dramatically, allowing water from the 60–90 centimeter layer and surface-derived water to percolate downward more freely. During the freezing period, bidirectional freezing — from the surface downward and from the base of the active layer upward — traps water above the advancing freezing front, yet some shallow water still finds its way to groundwater through preferential channels before the profile seals completely.

Isotopic evidence pointed to a subtle but important shift in recharge mechanics over the season. The researchers used a metric called line-conditioned excess, or lc-excess, which measures how far a water sample deviates from the local meteoric water line — the reference relationship between hydrogen and oxygen isotopes in precipitation. Water that has undergone evaporation, typical of slow, matrix-driven piston flow through soil pores, carries low lc-excess values, whereas water reaching the aquifer via fast preferential pathways retains an isotopic signature close to that of precipitation. Over the course of the freeze–thaw cycle, groundwater lc-excess values shifted from resembling soil water toward resembling precipitation, indicating that piston flow gradually weakened while preferential flow intensified, producing a recharge regime in which the two mechanisms coexist.

Space mattered as much as time. In the middle and upper reaches of the basin, where permafrost prevails, groundwater is recharged primarily by water from the 0–60 centimeter soil layer traveling along longer, more circuitous hydrological pathways. In the downstream regions dominated by seasonal frozen ground, by contrast, gentle topography and shallow water tables favor rapid infiltration, and groundwater is recharged mainly by water from the 30–90 centimeter layer moving quickly through macropores and fissures. Soil moisture measurements reinforced this picture: water content in the upper reaches remained consistently higher than downstream, and the seasonal patterns of wetting and drying differed markedly between the two zones, reflecting the different thermal regimes of permafrost and seasonally frozen ground.

The authors are candid about the limitations of their approach. Snowmelt samples were collected from the snowpack, so sublimation before melt and evaporation afterward may have altered the isotopic signature used to trace snow contributions. Freezing itself causes isotopic fractionation, as lighter isotopes preferentially remain in unfrozen liquid water while heavier isotopes enrich the ice phase, complicating pathway identification. And because piston flow and preferential flow operate simultaneously, water mixing between soil layers can blur the source signatures that the mixing model relies upon. Future work, they suggest, should combine high-frequency in situ monitoring with multiple tracers, including dye tracing, to pin down recharge pathways with greater precision.

Even with those caveats, the implications are far-reaching. As climate warming deepens the active layer and degrades permafrost across the plateau, the balance between recharge sources is likely to narrow while the differences between recharge pathways widen, reorganizing how water moves through these landscapes. The study demonstrates that assessments focused solely on rainfall risk underestimating the contributions of snowmelt and, above all, soil water to alpine aquifers. For water managers charged with sustaining baseflow in plateau rivers and the communities and ecosystems that depend on them, the message is clear: in frozen ground, the seasons do not merely change the weather — they rewrite the hidden architecture of the water cycle itself.

Subject of Research: Influence of freeze–thaw processes on shallow groundwater recharge sources and pathways in the Qinghai Lake Basin, traced with stable water isotopes

Article Title: Freeze–thaw processes influence shallow groundwater recharge sources and pathways in the Qinghai Lake Basin: insights from water isotopes

Article References: Zhang, W., Li, X., Deng, Y., Hu, G., & Shi, F. (2026). Freeze–thaw processes influence shallow groundwater recharge sources and pathways in the Qinghai Lake Basin: insights from water isotopes. Hydrology and Earth System Sciences, 30(18), 6039-6055. https://doi.org/10.5194/hess-30-6039-2026

Image Credits: AI Generated

DOI: 10.5194/hess-30-6039-2026

Keywords: groundwater recharge, freeze–thaw cycle, permafrost, Qinghai Lake Basin, stable water isotopes, Qinghai–Tibet Plateau, soil water, snowmelt, piston flow, preferential flow, active layer, alpine hydrology

The research, conducted by Gaojian Di of the Heilongjiang Province Hydraulic Research Institute and Bo Feng of the Changchun Institute of Technology, focused on the frozen soil region of Northeast China, an area where discontinuous permafrost and seasonally frozen ground interact with a warming boreal climate. Rather than treating the soil column as a single unit, the team examined four distinct soil layers, from the surface down to deeper horizons, capturing the depth-dependent character of climate signals as they propagate downward through the ground. This layered perspective matters because the surface responds almost immediately to atmospheric conditions, while deeper layers integrate and lag behind those signals, sometimes in surprising ways.

To build their analysis, the researchers drew on monthly ERA5-Land reanalysis data spanning 1997 to 2021, a dataset produced by the Copernicus Climate Change Service that blends satellite observations with numerical weather modeling to provide consistent estimates of land-surface conditions. Reanalysis products like ERA5-Land are invaluable in remote, cold regions where ground-based instrumentation is sparse, offering continuous records of soil temperature, soil moisture, radiation, precipitation, snow depth, and runoff across the entire study area. The team analyzed the spatiotemporal evolution of soil temperature and moisture across the four layers, identifying trends and patterns of interannual variability that would be difficult to detect from scattered field stations alone.

The headline result on temperature is strikingly orderly: soils warmed across all four layers during the study period, but the magnitude of that warming generally decreased with depth. This attenuation is physically intuitive, since heat propagating downward from the surface is damped by the thermal inertia of overlying soil, yet documenting the pattern quantitatively across a quarter century provides a benchmark for models of permafrost thermal evolution. The finding aligns with global observations that permafrost is warming worldwide, and it confirms that the frozen ground of Northeast China is following the same trajectory, with surface layers absorbing the brunt of atmospheric warming while deeper horizons warm more slowly but persistently.

Soil moisture told a messier story. Unlike temperature, moisture changes across the four layers displayed a non-monotonic pattern, meaning the direction and magnitude of change did not vary consistently with depth, and the records were accompanied by substantial interannual variability. This complexity reflects the tangled hydrology of frozen ground, where snowmelt timing, thaw depth, infiltration capacity, and evaporation all compete to determine how much liquid water resides in the soil at any given moment. In permafrost regions, the frozen matrix itself impedes drainage, so even modest shifts in thaw duration or precipitation phase can produce outsized and erratic swings in moisture content, complicating both ecological forecasting and engineering planning.

The methodological heart of the study lies in its use of XGBoost, a gradient-boosted decision tree algorithm renowned for its predictive power, paired with SHAP, or SHapley Additive exPlanations, a technique borrowed from game theory that quantifies each input variable’s contribution to a model’s output. Traditional machine learning models are often criticized as black boxes, but the XGBoost-SHAP framework allows researchers to open that box and see precisely which climatic variables explain soil changes at each depth, and how those contributions shift across space and time. This interpretable approach transforms the model from a mere curve-fitting exercise into a scientific instrument for hypothesis testing about the physical drivers of frozen soil dynamics.

For soil temperature, the attribution results revealed a clear depth-dependent hierarchy of drivers. Downward longwave radiation, the infrared energy emitted by the atmosphere toward the surface, made the largest contribution to explaining temperature variations in the surface and shallow soil layers. This makes physical sense: as atmospheric humidity and cloud cover increase under warming, they trap and re-emit more longwave radiation, directly heating the ground surface. For mid-layer soil temperature, however, snow depth emerged as the most significant explanatory variable, highlighting the dual role of snow as both an insulating blanket that shields the ground from extreme winter cold and a reflective surface that moderates spring warming. In the deepest layer, temperature changes were associated with the combined contributions of relative humidity, net surface shortwave radiation, and snow depth, suggesting that deep soil integrates multiple atmospheric pathways rather than responding to any single dominant driver.

The moisture attribution produced an equally compelling finding: total runoff demonstrated a high explanatory contribution across all four soil layers, outranking every other climatic variable. This underscores the tight coupling between surface hydrology and subsurface water storage in frozen terrain, where the generation and routing of runoff govern how water is distributed through the soil column. Following runoff, snow depth, surface air temperature, and total precipitation ranked as the leading contributors to soil moisture variation. The prominence of snow depth in both the temperature and moisture analyses reinforces its status as a master variable in cold-region hydrology, controlling not only the thermal insulation of the ground but also the timing and magnitude of the spring water pulse that recharges soil moisture as thaw proceeds.

Why should readers far from Northeast China care about the thermal state of its frozen ground? The reasons are both global and practical. Frozen soils store vast quantities of organic carbon, and their thaw can release greenhouse gases that amplify warming, a feedback loop of planetary consequence. Regionally, permafrost degradation alters groundwater flow, destabilizes infrastructure such as roads, pipelines, and buildings founded on ice-rich ground, and reshapes ecosystems that boreal forests and wetlands depend upon. Previous studies in the region have documented shrinking permafrost extent, thickening active layers, and shifting hydrological regimes, and the new quantitative attribution adds a crucial piece: it identifies which specific climatic levers are pulling hardest at each depth, information that can sharpen the projections of land surface and permafrost models.

The study also demonstrates a template for future cold-region research. By making the ERA5-Land dataset, the analysis code, and permafrost zoning data publicly available through the Copernicus Climate Data Store and a GitHub repository, the authors have lowered the barrier for other teams to replicate, extend, or challenge their findings. As machine learning continues to permeate the geosciences, the pairing of powerful algorithms with interpretability tools like SHAP offers a way to extract causal insight from big environmental data without sacrificing scientific transparency. For the frozen soils of Northeast China, the message is unambiguous: warming is penetrating the ground layer by layer, snow and radiation are the dominant conductors of that heat, and runoff holds the keys to the region’s soil water future. Understanding those mechanisms in detail is an essential step toward anticipating what a warmer century will do to the ground beneath our feet.

Subject of Research: Depth-dependent responses of frozen soil temperature and moisture to climate change in Northeast China, analyzed with XGBoost-SHAP machine learning

Article Title: Response of multi-layer soil temperature and moisture to climate change in the frozen soil Region of Northeast China based on interpretable machine learning

Article References: Di, G., & Feng, B. (2026). Response of multi-layer soil temperature and moisture to climate change in the frozen soil Region of Northeast China based on interpretable machine learning. Theoretical and Applied Climatology, 157(10), Article 658. https://doi.org/10.1007/s00704-026-06584-2

Image Credits: AI Generated

DOI: 10.1007/s00704-026-06584-2

Keywords: frozen soil, permafrost, Northeast China, soil temperature, soil moisture, climate change, XGBoost, SHAP, interpretable machine learning, ERA5-Land, snow depth, runoff