Different local gravimetric geoid solutions were carried out in Hungary
and the surrounding continental area. These solutions were based on terrestrial
gravity data and height data using as a reference surface the EGM96 geopotential
model. The gravity data used in the area 45.5° < Φ < 49°, 16° < λ < 23° were finally
gridded on a 1.5' x 2.5' geographical grid. These included more dense gravity data
with respect to a previous gravimetric solution for Romania and Yugoslavia. The
height data were available on a 1km x 1km grid. The methods used were the
spherical 1D Fast Fourier Transform (FFT) method and the Fast Co11ocation
(FCOL) procedure. In order to assess the accuracy of the computed geoid heights we
compared them with 43 Hungarian GPS/leveling stations belonging to the
EUREF89 network. The statistical results of the derived differences give an
accuracy close to 12 cm in terms of standard deviation, which decreases to 8 cm
after subtracting a linear trend and bias model. Excluding 14 GPS stations located
at the borders of Hungary the aforementioned accuracies reached the level of 7 cm
and 6 cm, respectively. Moreover, we have compared our geoid solution in the entire
test area with the European geoid EGG97 and found a standard- deviation of
differences close to 41 cm and 20 cm before and after subtracting a linear trend and
bias model. In a last numerical experiment we computed geoid heights by 1D FFT
using different cap sizes for the gravity data. The geoid height results were similar
and no significant improvement has been achieved.