The following post, written by Dr. Andriy Herts, describes how the Our Sci Reflectometer is being deployed in Ukraine to research and monitor soil contamination in Ukraine.

Over the past several years, our team from Ternopil Volodymyr Hnatiuk National Pedagogical University (Ukraine) has been testing the Our Sci Reflectometer across three separate research problems.

The team’s earliest published work with the Reflectometer answered a simple question: could a portable, fixed-wavelength device replace a full laboratory carbon analysis — at least well enough to be useful. Paired with a portable colorimeter, the device’s reflectance readings were regressed against lab-measured soil organic carbon, with one spectral range (500–632 nm) doing most of the work. The resulting model explained 69% of the variance in organic carbon content – not a replacement for the lab, but a fully legitimate first-pass screen, and proof that ten fixed wavelengths can carry real signal (Herts et al. 2022).

This result mattered less for the number itself than for what it opened up: a low-cost device that a research team without stable lab access could genuinely rely on, season after season, for the questions that matter most.

The second project applied the same logic to a more complex, higher-risk system. Returning to the soil the ash left over from burning Miscanthus × giganteus — a perennial energy crop grown on marginal land. Ash returns real nutrients — potassium, phosphorus, calcium – but exceeds the dose, and that same ash raises salinity and alkalinity enough to harm the very crop it was meant to feed.

Rather than run a full chemical analysis every time, the team scanned the soil with the Reflectometer across two growing seasons and found that a single near-infrared channel, 940 nm, captured the whole story. Reflectance at this wavelength decreased cleanly and steadily as ash dose increased, and correlated strongly with salinity (r = −0.81), available phosphorus (r = −0.85), and potassium (r = −0.83) — closely enough that a linear model built from 940 nm alone predicted phosphorus content with R² = 0.76. Plant development tracked the same signal: the healthiest growth fell within a specific reflectance range, and the values associated with plant stress were just as clearly visible. A two-year circular-economy experiment relied on a single wavelength that carried a double role — both a nutrient proxy and an early-warning system (Herts et al.,  in review; the study was carried out within the NATO Science for Peace and Security Programme grant G6094).

The last project (still ongoing), the one closest to the ground in a literal sense, is a pilot study the team designed around a much harder question: can that same low-cost device detect the layered damage in soil that has been both contaminated with diesel fuel and treated with biochar — a type of soil increasingly common across disturbed and post-military territories in Ukraine.

Over the first 6 months (under model conditions; the study is still ongoing, with a total duration of 12 months), the team tracked reflectance across eleven combinations of biochar and diesel doses. Biochar showed up immediately and remained stable: a strong, persistent negative correlation with NIR reflectance (ρ ≈ −0.86) that barely changed over half a year – consistent with biochar’s nature as a broadband light absorber. Diesel turned out to be a harder problem – the device’s range doesn’t reach the wavelengths where hydrocarbons show up directly – but it too left a real, reproducible trace: a dose-dependent “lightening” effect that intensified at longer wavelengths, evidence of a surface scattering effect rather than a color change. More tellingly – this relationship changed sign over six months, and at the highest biochar dose, the diesel effect peaked early and then declined – a pattern consistent with biochar likely physically binding the fuel within its pore structure over time, performing exactly the remediation work it’s meant for.

Fig. Mean Reflectance level of diesel effect across biochar levels

Of course, we recognize that none of this replaces a full laboratory hydrocarbon analysis, and we openly acknowledge that. Instead, we hope this will provide a fast, low-cost way to flag which plots need closer laboratory analysis – in conditions where making that decision quickly is not a matter of convenience, but of necessity.

Across all three projects, the throughline isn’t really about spectroscopy. It’s about whether a simple device can carry real diagnostic weight exactly where that matters most: in a circular-economy system that needed monitoring nobody had budgeted for, and in soil recovering from war, where the alternative to a fast field result is often not a slower lab result – but no result at all.

References

Herts, A., Khomenchuk, V., Kononchuk, O., Herts, N., Markiv, V., & Buianovskyi, A. 2022. “Use of visual-diagnostic color parameters of soils and optical reflectometry for determination of organic carbon content.” Journal of Geology, Geography and Geoecology 31 (2): 260–272. https://geology-dnu.dp.ua/index.php/GG/article/view/913

Herts, A., Kononchuk, O., Herts, N., Rouhani, A., & Pidlisnyuk, V. In review. “Rapid Vis–NIR assessment of residual soil transformation and plant response in a circular Miscanthus × giganteus biomass ash reuse system.”