Future dual-frequency, multi-constellation advanced receiver autonomous integrity monitoring (ARAIM) is expected to bring significant navigation performance improvement to civil aviation. The ARAIM user algorithm, which includes fault detection and exclusion (FDE) functions, is autonomously executed at the airborne receiver. To achieve specific integrity and continuity requirements, the real-time FDE process requires assertions on the signal-in-space (SIS) performance, and this information is carried in the integrity support message (ISM). This paper describes the design, analysis, and evaluation of the offline ground monitor, which aims at validating the ISM broadcast to users. To achieve this, GNSS satellite orbits and clocks must be estimated. There are many sophisticated orbit determination processes such as the one used by the international GNSS service (IGS), whose performance is specified in terms of accuracy. In contrast, the proposed offline ARAIM architecture is mainly intended for safety-critical aviation applications, in which integrity is of the primary concern. This monitor employs a straightforward approach to estimate satellite orbit/clock, which aims at facilitating ISM generation and validation. It takes advantage of the existing satellite based augmentation system (SBAS) ground infrastructure. In this paper, a worldwide network of sparsely distributed reference stations is considered, and parametric satellite orbital models are employed in the estimators, whose derivation and implementation are described step by step. Two separate analyses, covariance analysis and model fidelity evaluation, are carried out to respectively assess the impact of measurement errors and of residual model errors on the monitor's estimated orbit/clock. We have investigated different orbit models (GPS legacy versus CNAV orbital model) and reference station clock models (quadratic model versus no model). The results indicate the standard deviation of the monitor's orbit/clock estimation error is on the order of 30 cm, which is adequate for SIS performance validation.