Understanding sea level changes at a regional scale is important for improving local sea level projections and coastal management planning. Sea level budget (SLB) estimates derived from the sum of observation of each component close for the global mean. The sum of steric and Gravity Recovery and Climate Experiment (GRACE) ocean mass contributions to sea level calculated from measurements does not match the spatial patterns of sea surface height trends from satellite altimetry at 1 • grid resolution over the period 2005-2015. We investigate potential drivers of this mismatch aggregating to subbasin regions and find that the steric plus GRACE ocean mass observations do not represent the small-scale features seen in the satellite altimetry. In addition, there are discrepancies with large variance apparent at the global and hemispheric scale. Thus, the SLB closure on the global scale to some extent represents a cancelation of errors. The SLB is also sensitive to the glacial isostatic adjustment correction for GRACE and to altimery orbital altitude. Discrepancies in the SLB are largest for the Indian-South Pacific Ocean region. Taking the spread of plausible sea level trends, the SLB closes at the ocean-basin scale (2 ) but with large spread of magnitude, one third or more of the trend signal. Using the most up-to-date observation products, our ocean-region SLB does not close everywhere, and consideration of systematic uncertainties diminishes what information can be gained from the SLB about sea level processes, quantifying contributions, and validating Earth observation systems.
Plain Language SummaryAn important check on the accuracy of our global measurement systems and understanding of the processes driving sea level rise is the sea level budget. This describes the comparison of measured sea surface height change from satellite radar altimetry with the sum of its component parts, due to density and mass changes. With over 11 years of very high quality density and mass observations at good spatial scale, the sum of the parts matches the total within some uncertainty at a global scale. If, however, we examine the sea level budget at the ocean basin scale, we find significant discrepancies that are difficult to explain. We investigate different processing and averaging methods and find the density measurements do not include small spatial scales that are recorded by sea surface height measurements. Rather than this mismatch averaging out over basin scales, which we would expect if the errors are random, it instead leads to differences in the ocean basin average. Also, there is a mismatch at the hemispheric and global scale, which we believe comes from the way the satellite measurements are processed.