土壤是许多生态系统服务供给依赖的重要自然资源，水力作用下的土壤侵蚀是全球土地退化的重要原因。防治土壤侵蚀、减少土地退化，推动土地退化中性等可持续发展目标的实现，已经成为重要的全球性议题。降雨和植被是土壤侵蚀变化的两个主导因素。在过去几十年间，全球气候发生了剧烈变化，极端降雨事件增加，地表植被覆盖也发生了显著变化。在全球降雨和植被变化背景下，分析全球土壤侵蚀的时空变化特征，量化降雨和植被的相对贡献，预估全球土壤侵蚀未来变化趋势，对制定适宜的侵蚀调控与管理措施，减缓全球土壤侵蚀有着重要意义。

本文基于修正的通用土壤流失方程（RUSLE），比较和发展了大尺度土壤侵蚀因子评估方法，分析了过去三十年全球土壤侵蚀时空动态；量化了降雨和植被对土壤侵蚀变化的相对贡献，并针对不同气候区探讨了降雨和植被变化对土壤侵蚀的影响；预估了未来2021-2100年全球土壤侵蚀变化趋势，提出了相应的全球土壤侵蚀减缓措施及管理建议。主要结论如下：

（1）过去三十年全球降雨及降雨侵蚀力表现出明显的增加趋势。全球降雨量上升主要是由于侵蚀性降雨增加，大雨对全球降雨趋势和年际变化的贡献最大。空间上，降雨量与降雨侵蚀力变化格局基本一致，全球约有60%的地区呈增加趋势，其中15%的地区显著增加；全球仅有约3%的地区呈显著减少趋势。过去三十年全球受极端降雨影响的区域比例不断增加，而极端降雨对降雨侵蚀力变化有显著影响，全球尺度上最大一日降雨侵蚀力对年降雨侵蚀力的贡献超过10%，干旱区的平均贡献则超过50%。

（2）过去三十年全球植被显著改善，植被土壤保持能力持续增加。全球约70%的地区NDVI变化趋势为增加，包括40%的地区为显著增加，但其中有21.1%的地区未来变化趋势可能会发生反转；全球仅有10%的地区NDVI显著减少。全球尺度上，除干旱半干旱区外，大部分地区植被土壤保持能力对降雨增多不敏感，但几乎所有地区降雨减少都会导致植被土壤保持能力降低。因而当降雨减少时，由于植被侵蚀调节能力减弱，土壤侵蚀仍有上升的风险。

（3）过去三十年降雨主导了全球土壤侵蚀加剧，但植被增加有效缓解了降雨的影响。过去三十年，全球土壤侵蚀量从108亿吨增加到122亿吨，增加了约13%，在非洲南部干旱半干旱区域尤为显著；降雨是全球大部分区域土壤侵蚀变化的主导因素，其中降雨主导土壤侵蚀增加的区域占66%，降雨主导土壤侵蚀减少的区域占34%；在降雨侵蚀力增加的区域中，大部分地区植被土壤保持能力持续提升，52%的地区植被生长完全抵消了降雨侵蚀力增加的影响。

（4）降雨和植被对土壤侵蚀的影响与降雨量的变化范围有关。降雨侵蚀力、植被土壤保持能力与降雨量的关系为非线性，随着降雨量增加，降雨侵蚀力缓慢增加后再快速增加，植被土壤保持能力则表现出先线性增加后不再变化的趋势。因此在降雨量增加的过程中，降雨对侵蚀的影响逐渐加强，植被的影响则逐渐减弱，侵蚀的主导因素也逐渐由植被转变为降雨，其主要原因是大雨和暴雨事件逐渐增多，对降雨侵蚀力增加有显著影响，而植被则对该变化不敏感。

（5）未来变化情景下极端降雨和耕地面积不断增加，2021-2100年全球土壤侵蚀变化趋势将表现为持续增加。21世纪末降雨变化将导致全球侵蚀量增加13-25%，且在SSP5-RCP8.5情景下，未来全球侵蚀量将在降雨主导下持续上升。与过去三十年植被变绿有效缓解降雨增加对侵蚀的影响不同，未来SSP2-RCP4.5情景下耕地面积不断增加，土地覆盖变化成为全球土壤侵蚀增加的主导因素；在此情景下，土地覆盖变化将导致全球侵蚀量增加67%。

（6）推进全球土地退化中性可持续发展目标的实现，未来需要重点关注经济最不发达地区的土地恢复与管理。基于土壤侵蚀脆弱性评估框架，全球可以分为风险型、敏感型型以及调节型三大类区域，相应土壤保持策略分别为植被措施为主、工程措施为主、耕作措施为主。基于土壤侵蚀视角分析表明，全球约60%的国家无法实现2030土地退化中性目标，在非洲的刚果民主共和国、肯尼亚和埃塞俄比亚尤为突出。为推进土地退化中性可持续发展目标的实现，未来需要重点关注经济最不发达地区的土地恢复与管理。

Soil is an important natural resource , and provide many ecosystem services. soil erosion, especially water soil erosion, is an important cause of global land degradation. Preventing soil erosion, reducing land degradation, and promoting the realization of soil related sustainable development goals such as land degradation neutrality have become important global issues. Precipitation and vegetation are the two most important factors affecting soil erosion change. In the past few decades, extreme rainfall events have increased due to a shifting climate, meanwhile human activities such as farmland, urban expansion, and vegetation restoration have also significantly affected the vegetation cover. In the context of global precipitation and vegetation change, it is important to analyze the spatio-temporal change of global soil erosion, quantify the relative contribution of rainfall and vegetation and predict its future trends, in order to have a comprehensive understanding of surface soil erosion processes and adapt arppropriate management strategies to mitigate the negatice effects of future global changes on soil erosion.

This thesis first analyzes the spatial-temporal changes of rainfall for global and different climate zones, then analyzes the vegetation dynamics for global and different vegetation zones; this thesis also assess the spatio-temporal dynamics of global soil erosion in the past 30 years and quantified the relative contribution of rainfall and vegetation; Finally, based on the model simulation of future daily rainfall and land cover change, this study predicted global soil erosion for the futur 21st. The main conclusions are as follows:

(1) The global rainfall and rainfall erosivity have shown significnatly increased in the past three decades. The increase of erosive rainfall has promoted the upward trend of global rainfall, in which heavy rain dominanted the global annual rainfall trends and its inter-annual variability. Globally, the spatial pattern of rainfall and rainfall erosivity change is similar, about 60% area showed an increasing trend, including 15% showed significant increase. Only 3% of the surface showed a significant decrease trend. In the past three decades, the proportion of earth surface affected by extreme rainfall has continued to increase, and extreme rainfall has a significant impact on rainfall erosivity change, the contribution of maximum daily rainfall erosivity to the annual rainfall erosivity is more than 10% in both arid or humid regions, and contribution is more than 50% in some arid regions such as Sahel.

(2) The global vegetation has improved significantly in the past three decades, and the soil conservation capacity of vegetation has continued to increase. Spatially, about 70% of the study area showed an increasing NDVI trends, including 40% showed a significant increase, but 21.1% of these regions may reverse the change trend in the future; only 10% of the study area showed significantly decreasing trend. Except for arid and semi-arid areas, in most climate regions, vegetation is not sensitive to the rainfall increase. However, the decrease in rainfall will lead to a decrease in the soil conservation capacity of vegetation in almost all climatic regions. Therefore, when rainfall is significantly decreased, there is a risk that vegetation degradation will lead to increased erosion.

(3) In the past three decades, rainfall has dominated the global soil erosion increase, but vegetation increaseing at the same time largely offset the rainfall stress. In the past three decades, global soil erosion increased from 108×10⁹ t to 122×10⁹ t, increased by about 13%. rainfall is the dominant factor of soil erosion change for areas (65%) at the global scale, and most of the effects are negative. In areas where rainfall erosivity has increased, vegetation and its soil retention capacity has increased correspondingly in 80% of the regions, including 52% of areas where vegetation is completely offseted the impact of rainfall

(4) The dominant role of rainfall and vegetation on soil erosion is related to the variation range of rainfall. Rainfall erosivity and vegetation soil retention capacity increase non-linearly with rainfall. The typical change is: as rainfall increases, rainfall erosivity first increases slowly and then quickly, and vegetation soil retention capacity changes from a positive correlation with rainfall to not being affected by rainfall. Therefore, as rainfall increases, the impact of rainfall on erosion gradually strengthened, and the impact of vegetation gradually weakened, the dominant factor gradually transitioned from vegetation to rainfall. The main reason is that as rainfall continues to increase, the events of heavy rainfall gradually increase, this has a significant impact on the rainfall erosivity increase, while the vegetation is not sensitive to this change.

(5) Affected by the future extreme rainfall and the increasing of cropland, global soil erosion will continue to increase until the end of this century. Compared with the beginning of the 21st century, rainfall changes will cause global erosion rate increased by 13-25% at the end of the 21st century. At the same time, unlike vegetation offset the effects of rainfall in the past three decades, the area of cropland will continue to increase in the future in SSP2-RCP4.5 scenario, and land cover changes will become the dominant factor for the furue global soil erosion change, By the end of the 21st century, erosion is projected to increase by about 67% due to land use in the scenario.

        (6) To promote the realization of the global land degradation neutrality sustainable development goal, the key is the least developed areas. Based on the soil erosion vulnerability assessment framework, the world can be divided into three types of areas: risk type, sensitive type and regulation type. The corresponding soil conservation strategies are vegetation, engineering measures, and farming. Analysis based on the perspective of soil erosion shows that about 60% of the world's countries can not achieve the 2030 land degradation neutral target, especially in the Democratic Republic of Congo, Kenya and Ethiopia in Africa. In order to promote the realization of land degradation neutrality sustainable development goals, the future needs to focus on land restoration and management in the least developed areas.