Not every dry year means degraded land. A new framework published in Catena on August 25 can distinguish persistent degradation from short-term climate fluctuations.

Land degradation remains a critical ecological challenge worldwide, undermining ecosystem services, food security, and sustainable development. The United Nations Convention to Combat Desertification (UNCCD) has established Land Degradation Neutrality (LDN) as a core Sustainable Development target, aiming to balance land losses with gains through sustainable land management and ecological restoration by 2030.

Conventional LDN assessments, however, rely on static comparisons with a fixed historical baseline. In arid regions with high climatic variability, this approach often conflates short-term climatic fluctuations with long-term degradation trends, making it difficult to pinpoint areas experiencing persistent degradation that most urgently require intervention.

To address this, a research team led by Prof. Alishir Kurban from the Xinjiang Institute of Ecology and Geography (XIEG) of the Chinese Academy of Sciences (CAS), with Anwar Eziz, a special research assistant at XIEG, as first author of the team's study, has developed a novel Temporal Frequency Analysis (TFA) framework.

Rather than asking whether a pixel has degraded relative to an arbitrary past benchmark, TFA quantifies the recurrence of land degradation or improvement trends across consecutive assessment periods, using this recurrence as an indicator of landscape process persistence.

The method centers on the Land Productivity Dynamics (LPD) sub-indicator — the most responsive component of SDG indicator 15.3.1 in dryland systems — and draws on MODIS NDVI data spanning 2001-2022. The framework employed a five-year moving window and a full-period median dynamic baseline across ten overlapping assessment periods for pixel-level calculation of Degradation Recurrence (R_deg), Stability Recurrence (R_stab), and Improvement Recurrence (R_imp).

Highly localized persistent degradation

When the researchers applied TFA to the dryland provinces of Dashoguz and Lebap in Turkmenistan, they found that both persistent degradation hotspots and improvement bright spots were highly localized, each covering less than 1% of the study area. The finding suggests that severe, persistent land degradation is concentrated in specific locations rather than being widespread across the region.

Further analysis revealed distinct degradation patterns among different land-use types. In Dashoguz, degradation hotspots clustered in areas of severe secondary salinization and around Sarykamysh Lake, closely linked to mineralized drainage discharge. In Lebap, hotspots were more scattered and associated with saline soils, abandoned irrigated cropland, and overgrazing. Persistent bright spots, by contrast, corresponded to successful land-management practices — including natural vegetation regeneration in Dashoguz and efficiently irrigated zones along the Amu Darya in Lebap.

The two provinces differed in ecosystem resilience: districts in Lebap recorded stable-productivity recurrence rates of 74-76%, compared with only 23-46% in Dashoguz. The largest share of land — areas with 30-70% recurrence — was neither stable nor persistently degraded but dynamically unstable, representing both the greatest vulnerability and the greatest opportunity for intervention.

At the protected-area level, the Gaplangyr desert reserve showed only 13.80% persistently stable area, indicating pronounced fragility, whereas the Amudarya riparian reserve maintained 63.30% stable area, reflecting a resilient, balanced ecosystem. Among irrigated croplands, stable conditions covered 49.64% in Lebap versus just 24.71% in Dashoguz, where the area of persistent improvement slightly exceeded that of degradation, suggesting a net positive trend.

By distinguishing recurrent degradation from transient climatic variability, the TFA framework transforms LDN from a retrospective reporting obligation into a dynamic, process-oriented diagnostic tool. It provides spatially explicit decision support for the LDN "Avoid–Reduce–Reverse" response hierarchy: degradation hotspots can be targeted for reversal, dynamically unstable areas can be prioritized for preventive measures, and stable areas can be protected over the long term. Successful practices identified in improvement bright spots can also be documented and scaled up.

According to the researchers, the framework is readily reproducible without proprietary software or large-scale field campaigns built on open-source tools and globally consistent datasets. It offers a scalable methodological foundation for land-degradation assessment and adaptive management across the world's drylands.

Furthermore, as the researchers note, it carries particular relevance for advancing sustainable land-resource use and LDN implementation in the arid regions of Central Asia, particularly in countries involved in the Belt and Road Initiative.

Conceptual framework of the Temporal Frequency Analysis (TFA) method. (Image by XIEG)