In Trimble Business Center, the quality analysis of a geodetic network adjustment is addressed using the well-known Chi-square test, which compares the a priori and a posteriori variance factors. Based on a null hypothesis implying equality between both factors (reference factor = 1), the Chi-square test passes, indicating a successful adjustment. In contrary cases, i.e., an alternative hypothesis where both factors are different (reference factor ≠ 1), the Chi-square test fails, which may indicate suspicious observations or outliers, poor conditioning of the mathematical model, underestimated or overestimated stochastic models, errors in control coordinates, problems in Datum definition, or calculation errors.

 

 

If the null hypothesis fails, Trimble Business Center incorporates a strategy for identifying outliers based on the TAU test, which determines whether an observation fits the model obtained through residual analysis. The test results can be reviewed in the "network adjustment report."

 

 

From the histogram of standardized residuals, where the vertical axes to the right and left of the central vertical axis represent the maximum TAU values, it is possible to identify values that exceed these limits and are defined as outliers.

 

 

Once outliers are addressed, it is possible to apply a strategy based on weighting the matrix representing the stochastic model of the observations, thus addressing overestimated or underestimated models.