The construction of ecological security pattern, incorporating both soil and water conservation functions and geological disaster susceptibility, is critically important for preserving regional ecological stability and establishing robust early warning systems for natural disasters. This investigation centered on Northern Xinjiang utilized an integrated methodology to identify and analyze key components of the regional ecological structure. Ecological sources were delineated through the application of the Revised Universal Soil Loss Equation (RUSLE) model, which estimates potential soil erosion, and Morphological Spatial Pattern Analysis (MSPA), which identifies core habitat areas based on land cover patterns. Geological disaster susceptibility was comprehensively assessed using the Random Forest (RF) machine learning model, and the resulting susceptibility map was effectively employed to calibrate and modify the baseline resistance surface, thereby enhancing its accuracy. Building upon this refined resistance surface, circuit theory was implemented to simulate ecological connectivity and precisely identify ecological corridors, pinch points-areas critical for maintaining connectivity-and barrier points that impede ecological flows. The key findings were as follows: (1) Twenty distinct ecological source areas were identified, encompassing a total area of 81,763.56 km², which accounted for 20.96% of the total study area.. (2) Geological disaster susceptibility was evaluated using ten factors, including lithology and distance to faults. The Receiver Operating Characteristic (ROC) curve analysis yielded an Area Under the Curve (AUC) value of 0.89, approaching 1. High spatial coincidence between the susceptibility assessment and historical disaster points confirmed the model's reliability for resistance surface modification. After modification, high-resistance values were predominantly distributed in the Ili River Valley, Ta cheng Prefecture, and the northern Altai Mountains. (3) The ecological network analysis identified 37 ecological corridors totaling 2,914.86 km in length. This network consists of 19 important corridors (424.65 km) vital for regional connectivity and 18 general corridors (2,490.21 km). Furthermore, 53 ecological pinch points and 6 significant ecological barrier points were pinpointed, highlighting priority areas for conservation and restoration efforts. (4) Synthesizing the findings on geological disaster susceptibility and ecological security patterns, a structured "Three-Zone" ecological restoration framework-comprising core, buffer, and coordination zones-was developed. Specific protection strategies and tailored management measures were proposed according to the unique ecological challenges and functional positioning of each zone. The outcomes of this research provide a scientifically-grounded foundation and practical decision-support tools for ecological conservation planning, territorial spatial optimization, and disaster risk reduction initiatives in Northern Xinjiang. The results provided significant insights for safeguarding ecological security and enhancing geological disaster prevention in Northern Xinjiang.