Land degradation is the persistent reduction in the capacity of the land to support human and other life on Earth (IPBES, 2018). This process jeopardizes the provision of ecosystem services. The Sustainable Development Goal (SDG) 15, ‘Life on Land’, includes efforts to sustainably manage and recover natural ecosystems and restore degraded land and soil. Under the umbrella of SDG 15, the United Nations Convention to Combat Desertification (UNCCD) has defined an indicator framework to monitor progress toward ‘land degradation neutrality’. We evaluated the performance of SDG 15.3.1, focusing on “…proportion of land that is degraded over the total land area” for the European Union (EU) using the TRENDS.EARTH software.

We assessed the impact of alternative datasets at different spatial resolutions and policy-relevant data sources for land cover (CORINE) and soil organic carbon (SOC) stock (LUCAS). Our hypothesis was that higher spatial resolution sub-indicators would better identify the total share of degraded land and provide a clearer picture of the extent of degraded land for the target period. Land productivity trajectories were adjusted using the Water Use Efficiency index that revealed the high share of improving land reported by the NDVI trends. Therefore, it is advisable to use always a climate correction to assess land productivity trends. Replacing default datasets with alternative sub-indicators allowed the detection of 25–40% more degraded areas. Additionally, the integration with a combined proxy of land degradation (soil erosion >10 Mg ha−1 yr−1, and SOC concentration <1%) identified an additional 50% land degradation and revealed that a large extent of the EU needs restoration measures.

The SDG 15.3.1 indicator is able to recognize a part of the ongoing potential LD issues and not always capturing the severely degraded land. The LP fluctuations can onset a series of related impacts in the short run. Areas where degradation features coincide with the degradation class of the SDG 15.3.1 indicator are both natural and anthropic such as the Po plain valley and south of Spain; additionally, LD can be attributed to prolonged droughts, soil sealing, which are processes having a relatively higher incidence than deforestation and mining. LC changes (around 6% based on CORINE LC changes between 2000–2018) and SOC stock variation have played a very limited role in the calculation of the SDG 15.3.1 indicator being directly correlated each other in the calculation. According to the six scenarios provided in this work, land under degradation ranges from 11,1% to 23.3%, the area under stable condition from 27% to 77.5%, whereas improving conditions are seen in 4.9% to 60% of the land. The inclusion of areas with SOC values less than 1%, and soil erosion by water greater than 10 Mg ha−1 yr−1 has added roughly 50% additional share of land regarded as degraded.

Continued efforts are needed to quantify specific associations between LD and the reduction of ecosystem services. The limitations of the indicator in the context of agricultural lands, will serve as valuable input for discussions on EU wide definitions of the LD baseline, and standards on how to assess future changes (reporting periods). Therefore, urgent action is needed to tune the SDG 15.3.1 calculation approach for the EU to find solutions that can be implemented on regional, and local scale and at different institutional levels, taking into account the stakeholders’ interests and the natural capitals. In compliance with the UN SDG framework on using Earth observation data, there is still the need to assess which threat affects the degraded areas (i.e., persistent loss of biodiversity, ecosystem functions and services). SOC stock decline and soil erosion by water in most EU Member States are major threats, not only for agricultural land but also for uncultivated land. We found that the SDG 15.3.1 class “degraded” had only 20% agreement with a “SOC + Erosion” degradation layer.

The SDG 15.3.1 indicator does not takes into account other soil degradation phenomena such as slope stability, soil salinization, diffuse pollution, or loss of biodiversity. In other words, the indicator does not reflect clearly the broad spectrum of LD threats. Policy decisions might be wrongly based on such outcomes. This mapping exercise can highlight the need to look into each potential LD hotspot and to propose dynamic adaptation strategies and provide a sound basis for evidence-based policies. With this work, we highlighted the importance of contextualizing trends in LC and LP as well as the importance of soil monitoring (SOC stock) at a detailed scale. The LDN achievements assessed at national level must be checked at the landscape level, and the LD hotspots need to be restored through integrated land use planning.

Causes and impact in the short-middle term are still a LD major subject in light of the climate extremes and the agricultural intensification. The big picture of LD at local scale can only be achieved when the SDG 15.3.1 indicator will be combined with other LD remote-sensing-based measurement, socioeconomic, and other environmental data such as meteorological measurements, soil surveys, LC changes, crop productivity, socioeconomic, and demographic data. The so called “convergence of evidences” approach will allow for a more comprehensive interpretation of observed LP dynamics for each LC class in terms of LD, stable conditions, or improvement over time. Further work is needed for the evaluation of the high-resolution earth observation data to capture field-level potential LD, the sealing buffer effect, and the SOC stock decrease due to climate and agriculture.