Fr. Nabila PUTRI
Tropospheric delay models for GNSS in Indonesia

18. November 2021, 10:00, Zoom

Abstract:
This study assesses the performance of VMF3 for GNSS in Indonesia and compare it with VMF1 and GMF. Observations from eleven stations of the InaCORS network in 2014 were analyzed using PPP technique in Bernese GNSS Software 5.2. The estimated horizontal coordinates from the three mapping functions are similar, whereas the biases in height are larger. Furthermore, using either VMF3 and VMF1 provides slightly better results than GMF. The biases of the estimated ZWD between VMF3-VMF1 are smaller than VMF3-GMF. The mean differences of SWD at 5° using VMF3 and GMF range between 1-13 cm, and 0.4-6.5 cm for VMF3 and VMF1 .
PWV are derived from the estimated ZWD using water vapor weighted mean temperature, Tm. Tm was determined using temperature and humidity data from ERA5 and compared with Tm from Bevis formula. Tm Bevis generally has a cold bias compared to Tm ERA5 of less than -2 K. Based on Tm and surface temperature Ts from ERA5 at 40 grid points, Tm and Ts in Indonesia are not linearly related, with an overall correlation coefficient of 0.37.
PWV is then determined at nine stations in 2014 using Tm Bevis and Tm ERA5. The effect of mismodelling Tm to PWV estimation can reach a maximum of 2-3%, resulting in an error of around 1-2 mm. PWV at the nine stations vary from 20-70 mm, with the annual mean of 44-56 mm and standard deviations of 4.5-10.7 mm. Mean PWV during the dry season are 38-56 mm, and increases to around 49-56 mm during the rainy season.