No abstract

The Department of Lands (DOL), Thailand, has adopted the Network-based Real-Time Kinematic (NRTK) Global Navigation Satellite System (GNSS) surveying technique using a Virtual Reference Station (VRS) to support cadastral surveys since 2011. Determining accurate coordinates of parcel boundary markers at building corners or near fences and walls is difficult because a GNSS range pole cannot be leveled with a circular bubble. This study aims to evaluate the performance of the receivers equipped with tilt sensors for horizontal and vertical positioning. Two types of tilt sensors used for evaluation were a magnetometer and micro-electro-mechanical system (MEMS) and an inertial measurement unit (IMU). Conducting the NRTK GNSS surveying tests was based on the pole tilt angles of 0°, 15°, 25°, 35°, and 45° from a plumb line in controlled and obstructed environments. The IMU-based tilt sensor had more advantage of accurately positioning over the MEMS sensor. The results showed that using the IMU, better than 4 cm horizontal positioning accuracy was achievable when the pole was tilted by 15° or less under non-multipath and open-sky conditions. The vertical accuracy was of a few centimeter levels and least sensitive to tilt angles using either type of sensor. However, none of the sensors precisely compensated for pole tilt in strong-multipath and complex environments, causing increased horizontal errors in decimeter levels.

The Royal Thai Survey Department and Chiangmai University developed the Thailand geoid model 2017 (TGM2017) with a 1x1 grid to support the transformation between Global Navigation satellite System (GNSS) ellipsoid heights and Kolak-1915 vertical datum orthometric heights.TGM2007 was based on Thailand gravimetric geoid model 2017 (THAIG17) and 299 GNSS ellipsoidal heights co-located with Kolak-1915 heights. All terrestrial gravity data used for geoid computation came from the new national gravity network, consisting of 87 absolute and 9,929 relative gravity stations at 10-25 km intervals, mostly along with existing roads. From 2016 to 2017, airborne gravity surveys were conducted at a 4,000m-flight altitude and 10km along-track spacing to acquire the gravity data over mountainous and inaccessible areas, including coastal and marine areas, at an estimated accuracy of 3.0 mGal. Long-wavelength geoid structure was controlled by the GOCE-EGM2008 combined model (GECO) and the Technical University of Denmark's global marine gravity model 2013 (DTU13). All gravity data were combined and downward, using leastsquares collocation with the residual terrain model reductions from a digital terrain elevation data level 2 (DTED2). THA17G was determined by multi-band spherical Fast Fourier Transform and converted to TGM2017 with the 38.2cm root-mean-square (rms) fit of 299 GNSS/leveling co-points and a mean offset of 37.0cm. This value represents the separation between Kolak-1915 and a global mean sea level. The evaluation of TGM2017 at 100 GNSS/leveling checkpoints shows the rms of 4.9cm, consequently leading to reliable orthometric heights at a 10-cm accuracy level or better.