The phase 2 north America land data assimilation system (NLDAS-2) products for modelling water storage displacements for plate boundary observatory GPS stations

Abstract

In this paper, we compare the efficiency of two models to estimate the surface displacements due to continentalwater storage (CWS) variations over continental North America. The first model, themonthlyNorthAmerica Land Data Assimilation System Phase 2 Noah (NLDAS- 2 Noah) is a model of CWS restricted to North America. The second data set, the Global Land Data Assimilation System Noah (GLDAS Noah), is global. To compare the models, we use coordinate time series from GPS stations within the Plate Boundary Observatory (PBO). We find that the NLDAS-2 Noah CWS estimates of vertical surface displacements are correlated with PBO height coordinate time series with an average correlation of 0.4. Of the selected 986 PBO stations, stations with their weighted root mean square (WRMS) reduced after removing the surface displacements predicted using NLDAS-2 Noah surface mass, account for 13%, 27% and 56% for the north, east and up components respectively. The highest reductions in scatter occur on coordinate time series from stations in the mountains. Comparing NLDAS-2 Noah to GLDAS Noah, we find that the NLDAS-2 Noah model reduces the horizontalWRMS more than GLDAS for 88% and 73% of the PBO stations in the North and East components. In addition, stations in the mountains of the northwest and southeast part of the NLDAS-2 Noah spatial coverage (25% of the total stations) have their vertical scatter reduced by more than 10%. Therefore,we conclude that the NLDAS-2 Noah model better estimates the CWS induced 3-D surface displacement for PBO GPS stations in continental North America. The reasons may due to the finer spatial resolution, the updated Noah model, together with the more accurate surface forcing data of the NLDAS-2 Noah model.