Zhu, Y.; Zhang, J., and Tang, Q., 2021. An innovative method for the spatial sampling analysis of sea surface temperature in the Pacific-Indian Oceans. Journal of Coastal Research, 37(5), 1053–1062. Coconut Creek (Florida), ISSN 0749-0208.
In this study, a pragmatic approach is presented for evaluating the spatial representativeness of the point-scale drifting buoy sea-surface temperature (SST) over the joined area of Asia and the Pacific-Indian Ocean with that help of the high-resolution satellite-derived SST at multipixel scales. The relative spatial sampling error (RSSE) and coefficient of sill (CS) are selected to investigate consistency between the drifting buoy SST and pixel mean value of satellite-derived SST and representativeness of drifting buoy SST at pixel scales of 25 km, 50 km, and 100 km, respectively. The results show that at the 25-km scale, the consistency between the drifting buoy SST and the satellite-derived SST is high, and the spatial heterogeneity within the pixel scale is not obvious, i.e. only two measurements with RSSE are larger than the critical value of all 43,748 measurements. But with the increase in the scale, the number of the drifting buoy SST, which is inconsistent with the satellite-derived SST, is obviously increased (the number of the inconsistent point measurement at the scale of 50 km increased to 42, and at the scale of 100 km it is 59), and the spatial heterogeneity is enhanced. This spatial variation along the direction of latitude change is more obvious. Combining spatial consistency and spatial heterogeneity, no point-scale measurements of the worst spatial representativeness occur at the 25-km pixel scale, in which RSSE and CS are above the critical value. Although drifting buoy SST keeps a high consistency with satellite-derived SST, there are measurements of obvious local variation within pixel scales, and this variation is more obvious along the direction of latitude change. The spatial representativeness of the drifting buoy SST is unstable. This method is helpful to remove the poor-quality reference points during the validation of satellite-derived products, taking point-scale measurements as reference.