This paper evaluates, compares, and discusses different methods for quality control in geodetic data analysis in the general scenario of correlated observations and multiple outliers. The investigated methods are the data snooping procedure, the statistical tests for multiple outliers, the recently proposed quasi-accurate detection of outliers method for correlated observations, the Danish method for correlated observations, the robust estimator for correlated observations based on bifactor equivalent weights, and the robust estimator for correlated observations based on a local sensitivity downweighting strategy. To evaluate these methods, outliers between 3σ and 9σ magnitude (positives and/or negatives) are randomly generated and added to some observations (σ being the respective standard deviation of the observation) in two different global navigation satellite system (GNSS) networks that contain correlated observations. For each network, 15 000 scenarios are performed, 5000 with one outlier, 5000 with two outliers, and 5000 with three outliers, using Monte–Carlo simulations. The investigated methods have advantages and limitations, and the discussions and conclusions about the experiments are accurately presented.

Survey Review