Gross-Errors Detection in the Shipborne Gravity Data Set for Africa Hussein Abd-Elmotaal Atef Makhloof Minia University, Faculty of Engineering, Minia 61111, Egypt abdelmotaal@lycos.com Abstract In the frame-work of the African Geoid Project, a huge set of about 1.2 millions of shipborne gravity data points were collected. The shipborne data are collected in routes which intersect each others and merge together at the oceans surrounding the African continent. A scheme of gross-error detection within the shipborne gravity data set has been established employing the least-squares prediction blunder detection-removal technique to compute an estimation of the gravity anomalies at the data points without using the data values. The process works in such a way that it eliminates the blunders having a difference between the estimated and measured gravity anomalies more than three times the standard deviation of the whole data set. Hence, the process is repeated iteratively and re-computes the estimated values till the standard deviation of the differences becomes smaller than 1 mgal. Blunders of about 11% have been eliminated from the shipborne data set for Africa by the proposed technique. Key words: gross-errors detection, shipborne, Africa, least-squares prediction. 1. Introduction Data validation may include statistical tests to validate the reliability of the reconciled values, by checking whether gross-errors exist in the set of measured or observed values. These tests can be for example (Narasimhan and Jordache, 1999; Nyrnes el al. 2005) • the chi square test (global test). • the individual test. If no gross-errors exist in the set of measured or observed values, then each penalty term in the objective function is a random variable that is normally distributed with mean equals to 0 and variance equals to 1. The detection of possible gross-errors using the differences between observed values and values estimated using Least-Squares Collocation LSC technique has been successfully applied on a number of data types (cf. Tscherning, 1991 a, 1991 b; El-Tokhey and Abd-Elmotaal, 1996; Albertella et al., 2000). In this investigation, the least-square prediction detection-removal technique has been applied to remove the blunders from the shipborne gravity data set of Africa. Geodetic Week, Essen, Germany, October 8–10, 2013