ES.4 LIMITATIONS
The conclusions and recommendations offered here represent sound engineering judgements that are backed by considerable analysis. The timeframe for this study required that certain approximations be made in lieu of comprehensive simulations. The study results are believed to be conservative; margins were applied in those areas where the models and/or data sources were limited. The following limitations should be noted：
a. All performance analyses were based on snapshot measurement error statistics for an array of distributed geographic locations sampled every 5 min throughout one repeat cycle of the GPS constellation (i.e., one sidereal day). While this approach is believed adequate to estimate aggregate performance, verification of performance should be based on higher fidelity trajectory simulation.
b. Full aircraft trajectory simulations were restricted to evaluating interference effects using typical landing conditions with an antenna pattern derived from limited data sources. The television interference model was necessarily based on a very small data set.
c. No data were available to characterize high-definition television interference levels at the GPS frequencies.
d. Although the receiver model used to support this study is believed to be a good representation of typical receivers, the study did not explicitly account for actual receiver performance differences that may exist among users.
e. GPS/WAAS performance estimates were based on making adjustments to models derived from NSTB data. No detailed simulation was constructed for this analysis.
f. The ionospheric scintillation model used for this study was simplified, but the model used is believed to conservatively bound reality.
g. Time-to-alert analyses could not be explicitly included within the simulation structure used for these studies. The augmentation system's ability to meet these requirements was based on evaluations of the system design constraints provided by current descriptions and specifications.
