The ionosphere is one of the largest and unpredictable sources of error in single frequency Satellite Based Augmentation System (SBAS). The performance of the SBAS is significantly affected by the ionospheric behavior over the region of system operation. Thus, an ionospheric delay estimation algorithm based on geo-statistical interpolation technique, known as kriging, has been proposed to improve the accuracy of the estimation of ionospheric vertical delays at Ionospheric Grid Points (IGPs). This paper examines the benefits of incorporating the kriging based estimation algorithm in the ionospheric correction and integrity algorithms for the Korean SBAS which will be deployed in future. For this purpose, using the post-processed ionospheric observations from GNSS reference stations in South Korea, we first derive the variogram model to define nominal ionospheric residual covariance. We also examine the performance of the algorithm by making a comparison on two days, typical quiet day and stormy day, between the resulting estimates based on the kriging technique and the estimates obtained from the standard planar fit. It is shown that the implementation of kriging produces approximately 18% and 30% improvements in Root Mean Square (RMS) fit residual on the quiet day and stormy day respectively.

Keywords: SBAS (satellite based augmentation system), ionospheric delay estimation, kriging