As global navigation satellite system (GNSS)stations are sparsely distributed in oceanic area, oceanic areas usually have lower precision than continental areas on a global ionosphere maps (GIM). On the other hand, space-borne observations like satellite altimetry (SA) and ionospheric radio occultation (IRO) have substantial dual-frequency observations in oceanic areas, which could be used for total electron content (TEC) retrieval. In this paper, the Jason-2 SA and Constellation Observing System for Meteorology, Ionosphere, and Climate (COSMIC) IRO products were used to assess the precision of IGS GIM products. Both the systematic biases and scaling factors between the international GNSS service (IGS) GIM TEC and space-borne TEC were calculated, and the statistical results show that the biases and the scaling factors obviously vary under different temporal-spatial conditions. This analysis shows that these differences are variable with diurnal and latitude factors, that is, the differences in biases during the day time are higher than those during the night time, and larger biases are experienced at lower latitude areas than at high latitude areas. The results also show that in the southern hemisphere middle-high latitude area and some other central oceanic areas, the space-borne TEC values are even higher than GIM TEC values. As the precision of space-borne TEC should be evenly distributed around different areas on Earth, it can be explain that the TEC in these areas is undervalued by the current GIM model, and the space-borne SA and IRO techniques could be used as complementary observations to improve the accuracy and reliability of TEC values in these areas. maps using the Kriging interpolation algorithm, and the new Polytechnic University of Catalonia (UPC) kriging GIM has a lower RMS of about 16% and 2%, respectively, in the calculated slant TEC (STEC) than the original UPC GIM and IGS GIM in the self-consistency test [4]. Mautz et al. (2005) presented the B-spline wavelets method to represent the spatial and temporal variations of GIM and to solve the problems in ionospheric modeling due to data gaps [5]. The inequality-constrained least square (ICLS) method was proposed by Zhang et al. (2013) to eliminate nonphysical negative values in ionosphere associate analysis centers (IAACs) GIM products [2]. Li et al. (2015) proposed an approach named Spherical Harmonic plus generalized Trigonometric Series functions (SHPTS) to improve the accuracy and resolution of GIM; the accuracy of the SHPTS-based GIM could achieve 2-6 TECU over the area without GNSS measurements [6]. Wang et al. (2016) used priori VTEC values calculated from the IRI 2012 model to replace the grid points of GIM with negative VTEC values [7]. However, such interpolation and background constraint methods could not fully reflect the real situation of the ionosphere in ocean areas due to a lack of GNSS observations.As satellite-based observations are not constrained by the Earth's surface patterns, researchers are trying to use satellite-based dua...