Multifractal Analysis Reveals Natural Soil Chemistry Baselines on the Tibetan Plateau

High on the northern edge of the Tibetan Plateau, where the arid Golmud region meets the geologically complex terrain of Qinghai, scientists have completed one of the most detailed attempts yet to distinguish what soil naturally contains from what humans might have added. A new study published in Environmental Geochemistry and Health has derived regional geochemical background values and upper geochemical thresholds for fifteen elements in the surface soils of Golmud, using a multifractal statistical framework applied to an unusually dense sampling campaign of 3,319 surface soil samples. The work, led by Chaoliang Zheng of the Fifth Institute of Geological and Exploration of Qinghai Province together with colleagues at Yangtze University, addresses a problem that has long frustrated environmental regulators in geologically enriched regions: how to judge whether elevated concentrations of metals in soil reflect genuine contamination or simply the chemistry of the rocks from which the soil formed.

The significance of the study lies in the peculiar geology of the Golmud area. Situated on the northern Tibetan Plateau at the transition between the Kunlun Mountains and the Qaidam Basin, the region is underlain by intensely deformed orogenic belts, mafic and ultramafic intrusions, and mineralized zones that include the Xiarihamu nickel-cobalt deposit, one of the most significant magmatic sulfide discoveries in China. Soils developing over such parent materials can inherit naturally high concentrations of nickel, chromium, cobalt, copper and related elements. When national soil screening standards, which were calibrated largely against average conditions elsewhere, are applied without modification, large tracts of land can be falsely flagged as contaminated, triggering unnecessary remediation demands and stigmatizing agricultural land that is in fact geochemically normal for its setting.

To tackle this problem, the research team turned to multifractal analysis, a mathematical approach rooted in the recognition that geochemical distributions rarely follow simple Gaussian statistics. Concentrations of trace elements in soil tend to be scale-dependent and spatially clustered, with extreme values governed by nonlinear geological processes such as hydrothermal fluid circulation, magmatic differentiation and mechanical weathering. Fractal and multifractal methods, pioneered in geochemistry during the 1990s by researchers such as Qiuming Cheng and colleagues, model the frequency and spatial arrangement of concentrations across successive segments of the distribution, allowing the separation of background populations from anomalous ones without arbitrary cutoffs. The Golmud study applied a multifractal accumulated concentration-area frequency technique, fitting concentration-number relationships in log-log space and identifying breakpoints that mark transitions between distinct geochemical populations.

From these fitted segments, the researchers extracted two key quantities for each element: a regional background value representing the typical concentration across undisturbed soils, and an upper geochemical threshold marking the limit above which concentrations become geochemically anomalous. Crucially, the second fitted concentration segment in the multifractal model carries a characteristic singularity exponent, D2, which quantifies how rapidly concentrations decay through that segment. Elements such as cobalt, nickel, vanadium and copper displayed lower D2 values, below 2, indicating slower spatial decay and a tendency toward diffuse enrichment across broad areas, a pattern consistent with the widespread mafic and ultramafic lithologies of the East Kunlun orogenic belt. Chromium, by contrast, exhibited a markedly higher D2 of 2.86, reflecting a steeper decay and a more sharply localized enrichment behavior, potentially linked to discrete chromite-bearing sources or localized parent materials.

The comparison with national standards produced results that are both reassuring and illuminating. For lead, zinc, chromium and nickel, the upper geochemical thresholds derived for Golmud differed from the agricultural soil screening values specified in China’s national standard GB 15618-2018 by less than 15 percent, suggesting that for these elements the national values remain broadly applicable even in this geologically enriched setting. For arsenic, however, the regional threshold exceeded the national screening value, meaning that soils naturally enriched in arsenic could be misclassified as contaminated under the current regulatory framework. Conversely, the thresholds for copper and mercury fell below the national values, indicating that concentrations well within the natural range in Golmud would already surpass the national screening level, a finding with direct implications for how monitoring priorities are set in the region.

Beyond the eight elements covered by the national screening standard, the study derived regional thresholds for seven additional elements for which no element-specific screening values exist in GB 15618-2018: cobalt, uranium, thorium, manganese, vanadium, boron and phosphorus. This expansion matters for practical land management. Cobalt and nickel, for example, are of particular interest around the Xiarihamu deposit and the broader East Kunlun metallogenic belt, where exploration and mining activity is intensifying. Vanadium and manganese are increasingly scrutinized as potentially toxic elements in their own right, and phosphorus and boron play dual roles as both nutrients and potential contaminants in agricultural contexts. Providing locally calibrated reference values for these elements gives soil assessors in Qinghai a defensible baseline against which future monitoring data can be interpreted, rather than forcing them to borrow thresholds from fundamentally different geological environments.

The methodological implications extend well beyond Golmud. Establishing geochemical background values has long been recognized as one of the thorniest problems in environmental geochemistry, with researchers debating for decades whether background can be calculated at all, and how best to account for the enormous natural variability in soil parent materials. Traditional approaches, including medians, geometric means, and iterative outlier removal, assume relatively homogeneous populations and can misrepresent the true distribution in geologically heterogeneous terrain. The multifractal approach sidesteps many of these limitations by explicitly modeling the scale-invariant structure of concentration data, making it particularly well suited to regions like the Tibetan Plateau, where abrupt geological transitions occur over short distances and freeze-thaw weathering, glacial transport and aeolian deposition mix materials from multiple sources into the surface soil layer.

The authors are careful to note the limits of what their thresholds can do. Because these values are derived purely from the statistical structure of concentration distributions, they do not incorporate toxicity, exposure pathways, land use or bioavailability. A soil exceeding the regional threshold is geochemically anomalous, not necessarily hazardous. The researchers therefore position the new thresholds as complements to, rather than replacements for, regulatory screening values: they provide the geochemical reference frame needed to interpret the regulatory numbers sensibly, flagging where natural enrichment is expected and where genuinely anomalous concentrations warrant closer investigation. This distinction is critical for fair land-use decisions in mining districts and alpine agricultural zones alike, where applying unmodified national standards could either overlook real risks or impose unwarranted restrictions on safe land.

Looking forward, the study offers a template that could be replicated across the vast and geologically diverse territory of western China and beyond. Regional geochemical mapping programs, including China’s national geochemical baselines initiative, have generated enormous datasets of soil chemistry, and multifractal techniques provide a principled way to convert such raw data into locally meaningful standards. As mining, agriculture and infrastructure development press further into the Tibetan Plateau, the demand for scientifically defensible, region-specific soil assessment tools will only grow. The Golmud study demonstrates that with dense sampling and the right mathematics, the natural geochemical signature of even the most complicated terrain can be quantified, giving regulators, explorers and land managers a common reference point grounded in the actual chemistry of the land itself.

Subject of Research: Derivation of regional geochemical background values and upper geochemical thresholds in surface soils of Golmud, Qinghai, using multifractal analysis

Article Title: Regional geochemical background values and upper geochemical thresholds in surface soils of Golmud, Qinghai: a multifractal analysis

Article References: Zheng, C., Liu, Y., Jia, Y., & Niu, Y. (2026). Regional geochemical background values and upper geochemical thresholds in surface soils of Golmud, Qinghai: a multifractal analysis. Environmental Geochemistry and Health, 48(15), Article 602. https://doi.org/10.1007/s10653-026-03500-5

Image Credits: AI Generated

DOI: 10.1007/s10653-026-03500-5

Keywords: geochemical background, multifractal analysis, soil geochemistry, Tibetan Plateau, Golmud, Qinghai, soil screening values, heavy metals, Environmental Geochemistry and Health, upper geochemical thresholds, soil quality assessment, mining area