The increase in ecological risks caused by human activities has become a global concern in recent years. The Landscape Ecological Risk Index based on the theory of landscape ecology is more suitable for assessing large-scale ecological risks. Assessing landscape ecological risks and the mechanisms by which humans directly or indirectly affect them will help to manage and control the regions’ ecological risks through scientific and policy methods. In this study, a new model of landscape ecological risk assessment based on the moving window method is proposed. The Loess Plateau of China is used as an example, and the Human Footprint Index dataset of the Loess Plateau is constructed. Different human footprint factors and climate factors are applied, and the human direct and indirect effects on the landscape ecological risks of the Loess Plateau are explored based on the geographical detector model. The results show that, in 2000, 2010 and 2020, the landscape ecological risks of the Loess Plateau are currently in an unstable state, and the highest value area of the Landscape Ecological Risk Index continues to expand, with values of 113,566.1553 km², 114,575.6772 km² and 120,718.5363 km², respectively. Among all the human footprint factors, the population density factor has the highest effect on the landscape ecological risks of the Loess Plateau. Among the climate factors, both the average temperature factor and the average lagged temperature factor have significant effects on the landscape ecological risks of the Loess Plateau. With the interaction of any two human footprint factors and climate factors, the effect of these factors on the landscape ecological risks of the Loess Plateau is enhanced.

The ecosystem service value (ESV) is a critical metric for assessing the construction and protection of the environment. The research into the ESV pattern and the future development trend in the Loess Plateau of Northern Shaanxi is important for the conservation of water and soil and the enhancement of the natural environment in the region. In this study, the variations and distribution patterns of the ESV in the study area from 2000 to 2020 were analyzed, the influence of various natural and social factors on the ESV was quantified, the weight of each factor was analyzed and evaluated using the entropy weighting method, and, finally, a prediction was made regarding how the ESV will develop going forward in this area. The results show that (1) the ESV showed a decreasing trend from 2000 to 2020, with the highest value for soil conservation and the lowest value for food production. Among the 25 districts and counties, Suide County had the lowest ESV per unit area, whereas Huanglong County had the highest. (2) The global positive correlation was clearly visible in the ESV. According to local spatial autocorrelation analysis, the area had a “high-high” agglomeration area in the south and a “low-low” agglomeration area in the middle and north. (3) Among the various influencing factors, population density had the highest weight and the distance from roads had the lowest weight. The impact status of the area generally showed a lighter impact in the southern region and a heavier impact in the northern region. (4) In 2030, the total ESV is predicted to be CNY 4343.6 million in the study area, CNY 39 million lower than that in 2020.