Determinação e Validação de Modelos de Separação de Superfícies com Referência ao Elipsoide pelo GNSS e GNSS/INS

Abstract

Advances in GNSS technology allowed greater accuracy in the positioning of hydrographic surveys, so that the variation in ellipsoidal heights of a vessel could be representative of the variation in the local tide. For this purpose, it is necessary to define a separation model (SEP) between the Chart Datum (CD) and the ellipsoid, which can be determined by combining sea level data from a tide gauge station with ellipsoidal height data from a hydrographic boat. In this work, such methodology will be evaluated with different GNSS positioning methods and coupling modes to the inertial sensor. As a validation criterion, an RMSE (CI 95%) of 1 m was adopted for the horizontal component and 5 cm for the vertical component, which were evaluated by two approaches: the first, with the boat moored next to a tide gauge and comparing , relatively, the variation of the vessel's ellipsoidal heights with the variation of the tide observed at the station; and the second, with the boat moored and sailing, the most accurate method was adopted as a reference: Inertial Aided Post Processed Kinematic – Tightly Coupled (IAPPK (TC)) to evaluate the vertical and horizontal components of the other methods. In the first approach, the GNSS data were resampled at the same rate using the Spline function. Hampel and moving average filters were used for removal, adaptation of outliers and curve smoothing. Its parameters were empirically defined, in order to minimize the RMSE. In all the four periods analyzed with the boat moored, the methods based on the PPK obtained an RMSE in relation to the tide gauge below 5 cm; methods based on RTK and PPP, in three periods; and the GcDGNSS, in any period. In the second approach, for the eleven periods analyzed with baselines smaller than 15 km, the methods based on PPK obtained vertical uncertainties smaller than 5 cm. With the IAPPP method (TC), this occurred in eight periods; with the IAPPP (Loosely Coupled - LC), in six; and with PPP methods based only on GNSS, in four. With regards to horizontal uncertainties, in all days and methods, a value lower than 30 cm was observed. After validating the geodetic methods, the SEP of the tide gauge station at Clube Naval Charitas was calculated by combining GNSS tracking data from a level reference with leveling data from the station's F-41 sheet. Subsequently, this result served to validate the SEP calculated by the hydrographic boat, moored close to the tide gauge, using the IAPPK (TC) method. In comparison with the SEP of the tide gauge station, a standard deviation of 5.3 cm was found, with a mean difference of 0.1 cm. That said, it was found that the methods based on the PPK are the most capable of defining the SEP within the vertical uncertainty criterion of 5 cm. Finally, it should be noted that this research is important for the validation of surfaces with reference to the ellipsoid (e.g geoid, mean sea level topography) and all applications related to maritime positioning.