The applicability of degree-variance propagation for error analysis an
d for the validation of other error prediction methods is investigated
for the case of satellite gradiometry. For several idealized mission
scenarios exact expressions for the a posteriori variances exist; for
certain other scenarios approximate relations exist. As degree varianc
es cannot be directly applied to a non-polar orbit, a rule of thumb to
predict which coefficients will be affected by the polar gap is deriv
ed.