Abstract: Assessing the health of regional farmland ecosystems from the perspective of uncertainty can improve the scientific basis and accuracy of agricultural economic decision-making. Based on the concept of ecosystem health, we constructed an evaluation index system of farmland ecosystem health considering four dimensions: carrying capacity, productivity, competitiveness, and innovation. Combined with Monte Carlo random simulation, we assessed the influence of weightsetting uncertainty on health evaluation results and revealed the key role of weight changes in the 10-year evaluation. The results showed that grain yield per unit area increased significantly (by 32.9%-56.0%) in Henan Province from 2011 to 2020, but farmland ecosystem health level varied markedly among regions. The Central (Zhengzhou), Southern (Nanyang), and Western (Luoyang) regions were mainly at the sub-healthy level due to high pressure on cultivated land resources, with membership probabilities of 46.3%, 59.3%, and 67.4%, respectively. In contrast, the Eastern (Kaifeng) and Northern (Anyang) regions experienced significant improvements, driven by increased grain yield and enhanced ecosystem productivity. However, insufficient innovation limited the overall health, which remained mainly sub-healthy, with membership probabilities ranging from 39.8% to 45.2%. Based on uncertainty simulation and spatiotemporal analysis, this study accurately identified the trends and main influencing factors of regional farmland ecosystem health levels, and provides theoretical support and data reference for improving the scientific basis and accuracy of agricultural economic decision-making.

Key words: agricultural economy, evaluation index system, Monte Carlo simulation, spatial characteristics, Henan Province