The aff’s use of models and empirical data surrounding soil erosion is key to uniform and effective policy strategies that get modeled on a global scale– that empowers action but data is key to strategic analysis

Soil erosion can be mitigated using sustainable land management techniques and suitable policy incentives. Arable land, subject to conservation agriculture (CA) worldwide, is estimated to cover 11 to 14% (current study) or ∼1.42 billion hectares globally (20). Compared to a baseline scenario without any soil conservation practices, estimates indicate an overall global soil erosion reduction of about 7.1% under conservation practices (12). The footprint that human activities have left on the world’s soils are tangible (3). Soil erosion also affects nutrient and soil carbon cycling (21). The associated losses of nutrients, such as nitrogen and phosphorus, and organic carbon (22) have shown compromising long-term effects on the local ability of certain soils to meet agricultural production and ecosystem service demands (9). Global consumption of the three main fertilizer nutrients (N, P, and K) used to maintain, or improve, soil fertility is currently growing on a yearly average by 1.5, 2.2, and 2.4%, respectively (23). This comes along with high onsite economic costs for the land users to contain the production losses (12) and important cascading offsite environmental impacts (24). The pressure on fertile soils (9) through an exacerbation of soil erosion and its environmental degradation effects (25) is further strained by 1) the growing population with an estimated peak of 9.4 billion in 2070 (26), 2) the trend toward meat-intensive diets, and 3) a global climate that tends toward a more vigorous hydrological cycle (27, 28) as discussed above. These are all issues policy teams need to deal with. Today, policy makers emphasize the need for an evidence-based approach. This modeling effort forms part of a wider program of work coordinated through the United Nations to inform national policy makers. Soil erosion is closely linked to gross domestic product (GDP) (12), with poorer countries often experiencing the greatest impacts. Hence this effort is designed to provide a globally consistent assessment to help target policy effort. It will inform countries, especially those that are often least able to identify risk, to focus resources to mitigate degradation by erosion most efficiently and effectively. Moreover, the UN endorsed the World Overview of Conservation Approaches and Technologies (WOCAT) database (29), which contains freely available information on sustainable land management (SLM) practices. By combining the model outputs with WOCAT SLMs, the two resources give decision makers a powerful set of national scale tools to identify water erosion hotspots at appropriate scales and mitigate them locally through intervention planning. While past studies have provided insights into the possible future trends of global soil erosion by water with reference to land use and climate changes (13), they have not offered the resolution for guiding policy intervention. Unique to our study is the attempt to project future changes in soil erosion at the global level using the state-of-the-knowledge harmonized set of land use and climate scenarios developed according to the new SSP-RCP adopted by the UN Intergovernmental Panel on Climate Change (IPCC). This article is a synthesis of UN collaborative effort to tackle this important socioeconomic and environmental challenge; contributing evidence in support of the upcoming UN Decade on Ecosystem Restoration (2021 to 2030). Our hope is that this contribution, along with others, will inform and empower decision makers globally to tackle soil threats and develop concerted global and national strategies for soil monitoring, conservation, and restoration. RESULTS AND DISCUSSION Achieving the UN Sustainable Development Goals (SDGs) and the UN strategies for soil conservation (30) requires understanding of the location and magnitude of erosion at global scales, now and in the future. Here, we address this challenge, in support of decision-making, by forecasting global changes in soil erosion by water, driven by land use and climate change, until 2070. We analyze mitigation opportunities through current standards of CAand reflect on possible socioenvironmental impacts related to future changes in water erosion. We also discuss the limits of our approach, quantifying the uncertainty of our estimates. To estimate future soil erosion rates, we feed the high-resolution (250 × 250 m) RUSLE-based (12) modeling platform Global Soil Erosion Modeling (GloSEM) (SI Appendix, Soil Erosion Modeling) with future projections of land use obtained from the integrated assessment model (IAM) (30) (SI Appendix, Global Land Use/Cover and Future Change) and climate erosivity computed following the approach used in previous continental and global scale studies (18, 31) (SI Appendix, Climate and Future Change). The 250 m pixel resolution reflects land cover at the field scale, capturing the emergent mosaic of land use that facilitates or mitigates erosion processes. Multiple land use and climate scenarios, following three Representative Concentration Pathways (IMAGE SSP1-RCP2.6, MESSAGE-GLOBIOM SSP2-RCP4.5, and REMIND-MAGPIE SSP5-RCP8.5, with the numbers referring to the increase in radiative forcing in W·m−2 by 2100 (SI Appendix, Representative Concentration Pathway) greenhouse gas concentration trajectories adopted by the IPCC for its Fifth Assessment Report (AR5) in 2014, are modeled. Despite some limits and uncertainties addressed in the course of the manuscript, the modeling approach allows for a first estimation of soil displacement by water erosion due to sheet and rill erosion processes (referred to as “soil erosion” in the following technical definition in SI Appendix, Definition of Soil Erosion). These are the dominant processes on upper hill slopes because they are responsible for the largest part of soil displacement on agricultural fields and intensively used grassland (12) and found to be a fair proxy of a wider set of soil erosion processes by water (32). What is the global pattern of soil erosion? The worldwide spatial pattern of estimated soil erosion in 2015 is illustrated in Fig. 1A. Our modeling results suggest that water erosion is a common phenomenon under all climatic conditions across all observed continents. However, the distribution of the spatial soil erosion patterns suggests that soil erosion seems to threaten areas of large-scale reclamation such as major agricultural sectors, especially if it occurs in conjunction with concentrated intense rainfall events (Southern Brazil, Argentina, India, East China, Midwestern United States, Ethiopia, and Mediterranean Europe). Locally, steep slopes and high-relief topography also experience high erosion rates (e.g., Andes, Himalayas, Verkhoyansk Range, and Alaska Mountain Range) together with regions with generally sparse vegetation cover across the year. In the 2015 scenario, we estimate global soil erosion equal to 43+9.2−7 Pg yr−1. The confidence intervals account for the uncertainty of the spatial predictions estimated using a Markov Chain Monte Carlo (MCMC) approach, the uncertainty of estimating the area under CA the uncertainty related to the effectiveness of the CA practices, and the uncertainty in regional rainfall intensity-kinetic energy relationships. The increase of about ca. 8 Pg yr−1 compared to previous figures reported in Borrelli et al. (12) is a consequence of the increase in the global modeling area (from ∼84.1 to ∼95.5% of the Earth’s land surface) and to a lesser extent to the land use differences derived from adopting the land use data of Hurtt et al. (30) in the current study. This estimate is lower than the figures previously presented in the scientific literature adopting similar modeling approaches, e.g., Yang et al. (13) (ca. 132 Pg yr−1) and Ito (14) (ca. 172 Pg yr−1). It supports the findings reported in the latest reference document (6) of the UN on the status of global soil resources, which indicates a value below 50 Pg yr−1 as a more realistic quantitative figure of global soil erosion by the processes considered (6). Insights in support of the plausibility of our modeling estimates and GloSEM limitations are reported below. We provide further information on Model Performance Evaluation and Limitations of GloSEM below and in the SI Appendix. How does land use impact our estimates? Comparing soil erosion rates according to land use types, we find a decline in the estimates from croplands to forests and other forms of vegetation. Soil erosion rates are in line with field measurements (7, 33). Annual crops covered about 16% of land in 2015 and are estimated to be responsible for 41% of the total predicted soil erosion. Overall, the main agricultural lands (annual crops, permanent crops, and managed pasture) are responsible for 54% (equal to 23.4+5.3−4.1 Pg yr−1) of the total soil erosion. Future scenarios suggest that the effects of land use change may either decrease [SSP1-RCP2.6 (Fig. 1B)] or increase [SSP2-RCP4.5 (Fig. 1C), SSP5-RCP8.5 (Fig. 1D)] soil erosion by 2070. The divergent trends are the result of the possible different human development and societal choices described in the three considered SSP-RCP. The IMAGE SSP1-RCP2.6 scenario, which represents a pathway aiming at limiting the increase of global mean temperature to a maximum of 2 °C by 2100, suggests a possible contraction of the main agricultural land by 2070. The simulation of this land use scenario in GloSEM yields a potential decrease in soil erosion by water of –10% (global soil erosion equal to 38.5+8.6−6.2). The decrease in the simulated global soil erosion (∼4.5 Pg yr−1) is the result of a global reorganization of future lands, which according to the IMAGE SSP1-RCP2.6, will tend toward an overall decrease of agricultural areas in favor of an increase in forest and seminatural vegetation areas (SI Appendix, Fig. S1A). The annual crops and managed pasture are estimated to decrease globally by ca. 0.9 and 2.7 million km2 (equal to −2.9 and −2.4 Pg yr−1 of soil erosion, respectively) under the SSP1-RCP2.6 scenario, whereas the permanent crops follow a different trend with a potential increase estimated at 1.3 million km2 (+0.2 Pg yr−1). This scenario indicates a possible reduction of the share of soil erosion in agricultural land from ∼54% in 2015 to ∼48% in 2070. By contrast, the share of soil erosion in agricultural land increases to 56% and 59% under the MESSAGE-GLOBIOM SSP2-RCP4.5 and REMIND-MAGPIE SSP5-RCP8.5 scenarios, respectively. The MESSAGE-GLOBIOM SSP2-RCP4.5 is a low stabilization scenario that stabilized radiative forcing at 4.5 W/m2 (∼650 ppm CO2-equivalent) before 2100. Under the SSP2-RCP4.5 scenario, soil erosion by water would experience a slightly increased estimate at +2% (global soil erosion equal to 43.9+9.1−6.8), mostly driven by the expansion of annual crops estimated at 2.1 million km2 (+2.6 Pg yr−1). This increase is partially compensated for by the resulting contraction of the managed pasture (−1.2 Pg yr−1) and nonagricultural lands (−0.5 Pg yr−1). Geographically, the SSP2-RCP4.5 shows some mixed trends (SI Appendix, Fig. S1B). Major drivers of the increase can be found in Sub-Saharan Africa, Eastern Europe, some parts of Eastern Asia, and South America. The last scenario considers very high greenhouse gases (GHG) emissions (REMIND-MAGPIE SSP5-RCP8.5) and yields a possible notable increase in future soil erosion by water of +10% (global soil erosion equal to 47.3+9.5−7.3). The overall increase of soil erosion by water is estimated to be ∼4.3 Pg yr−1, primarily associated with substantial increases in agricultural areas in Sub-Saharan Africa, Brazil, India, Myanmar, and some districts of China. The land use conditions reported by the SSP5-RCP8.5 scenario (SI Appendix, Fig. S1C) for sizable agricultural districts in North America, Europe, and Russia would result in a lack of erosion reduction that characterizes the scenario SSP1-RCP2.6, and to a lesser extent in the SSP2-RCP4.5 scenario. What is the combined effect of future land use and climate projections? The modeling results (Fig. 2 A and B) suggest that climate change is the major driver of the change in soil erosion. The combined land use and climate simulations show a potential substantial increase in average soil erosion totaling +30% (SSP1-RCP2.6), +51% (SSP2-RCP4.5), and +66% (SSP5-RCP8.5). Quantitatively, 56.1+20.6−16.4, 64.8+28.5−21.4, and 71.6+32.5−24.7 Pg yr−1 are predicted for the SSP1-RCP2.6, SSP2-RCP4.5, and SSP5-RCP8.5 scenarios. The wider confidence intervals are related to the error propagation accounting for the variability of future climate projections of the 14 GCMs used to assess future rainfall erosivity (Fig. 3) and the set of uncertainties already considered in the 2015 scenario. These large confidence intervals associated with future climate projections reflect the higher uncertainty of these estimates. The multiscenario comparison suggests that although future land use changes can notably affect global soil erosion processes through the expansion or contraction of croplands, a global climate potentially moving toward more vigorous hydrological cycles would be acting as a major driver of future increases in soil erosion (Fig. 4). All SSP-RCP climate scenarios processed in the Gaussian process regression (GPR) model (SI Appendix, Climate and Future Change) indicate possible substantial increases in future rainfall erosivity following similar spatial patterns but different intensities (Fig. 2 and SI Appendix, Fig. S2). In absolute terms, the greatest increases occur in areas with tropical climates. Countries in temperate latitudes and with subtropical climates, however, would not be spared by increases of climate erosivity and experience increase peaks up to 50%. In these climatic zones, substantial increases might interest sectors of Eastern North America, Central and Northern Europe, Middle East, and North and East Asia. This provides important insight for policy makers identifying hotspots and trying to mitigate land degradation around the globe. What are the key findings for policy makers? Article 1 of the UN Convention to Combat Desertification (UNCCD), ratified in March 2020 by 197 countries, identifies soil erosion as a primary cause of land degradation, which, in turn, contributes to poverty and inequality through its negative effects on agriculture, food security, and ecosystems (34). Scientific evidence suggests that global warming has already increased global economic inequality (35) and influenced immigration waves (36). The remarkable increases predicted in the combined future land use and climate change scenarios could have a wide range of negative effects at global, regional, and country levels for which this strategic analysis is intended, and not the field or catchment scale. The first and most obvious consideration emerging from this analysis is the possible exacerbation of the climatic conditions, which could substantially increase global soil erosion (+30%) already in the low GHG emission scenario (SSP1-RCP2.6). A climate-induced increase in soil erosion, as indicated by our findings, with associated land degradation and loss of key ecosystem services (9), would pose a serious threat to the achievement of a large set of targets defined by the UN strategy with the SDGs. This is particularly true for achieving 1) a land degradation neutral world by 2030 (Goal 15; Target 15.3); 2) maintaining soil quality for achieving food security (Goal 2); 3) ensuring availability and sustainable management of water resources (Goal 6); as well as 4) ending poverty (SDG 1); 5) reducing inequality (SDG 10); and 6) taking action to combat global warming (SDG 13).