The advancement of autonomous multi-robot systems for planetary exploration missions is critically dependent on their ability to navigate accurately in unstructured and challenging environments, a primary source of localization failure in such contexts is the significant error introduced by unmodeled vehicle dynamics, particularly terrain-induced lateral slip, which severely degrades the performance of standard Simultaneous Localization and Mapping (SLAM) algorithms, this paper presents a terrain-aware cooperative SLAM framework designed to enhance localization accuracy for multi-robot teams operating on undulating surfaces, the proposed approach utilizes an Extended Kalman Filter (EKF) that uniquely integrates inter-robot relative measurements, derived from LiDAR, with a sophisticated motion model that explicitly predicts and compensates for lateral slip based on a 2.5D terrain map, the system's performance is rigorously evaluated within a high-fidelity simulation environment featuring complex topography, realistic sensor and motion noise, and rudimentary communication constraints. Quantitative and qualitative results demonstrate that the proposed framework significantly improves pose estimation accuracy and maintains robust performance over long-term operations, this work validates the efficacy of combining cooperative information sharing with explicit modeling of terrain-motion interaction as a vital strategy for enabling robust multi-robot exploration.