The lack of direct measurement of root-zone soil moisture poses a challengeto the large-scale prediction of ecosystem response to variation in soilwater. Microwave remote sensing capability is limited to measuring moisturecontent in the uppermost few centimetres of soil. The GRACE (Gravity Recoveryand Climate Experiment) mission detected the variability in storage withinthe total water column. However, root-zone soil moisture cannot be separatedfrom GRACE-observed total water storage anomalies without ancillaryinformation on surface water and groundwater changes. In this study, GRACEtotal water storage anomalies and SMOS near-surface soil moistureobservations were jointly assimilated into a hydrological model globally tobetter estimate the impact of changes in root-zone soil moisture onvegetation vigour. Overall, the accuracy of root-zone soil moisture estimatesthrough the joint assimilation of surface soil moisture and total waterstorage retrievals showed improved consistency with ground-based soilmoisture measurements and satellite-observed greenness when compared toopen-loop estimates (i.e. without assimilation). For example, the correlationbetween modelled and in situ measurements of root-zone moisture increasedby 0.1 (from 0.48 to 0.58) and 0.12 (from 0.53 to 0.65) on average forgrasslands and croplands, respectively. Improved correlations were foundbetween vegetation greenness and soil water storage on both seasonalvariability and anomalies over water-limited regions. Joint assimilationresults show a more severe deficit in soil water anomalies in easternAustralia, southern India and eastern Brazil over the period of 2010 to 2016than the open-loop, consistent with the satellite-observed vegetationgreenness anomalies. The assimilation of satellite-observed water contentcontributes to more accurate knowledge of soil water availability, providingnew insights for monitoring hidden water stress and vegetation conditions.
展开▼
