The AAU Calibration and Data Assimilation (CDA) approach for improving large-scale hydrological models in a changing climate

Extreme events such as floods and droughts are expected to become more frequent and intense due to the changing climate. However, it is still a challenge to monitor, understand, simulate, and anticipate the underlying hydrological processes. Large scale hydrological models and remote sensing observations (such as surface soil moisture from SMOS, SMAP and Sentinel, as well as total water storage changes (TWSC) from the GRACE and GRACE-FO gravity missions) provide a unique large-scale to global view on the changing hydrology. Sequential Calibration and Data Assimilation (CDA) provides opportunities to combine benefits from both, modelling and observing, and thus helps to improve our understanding of the impact of the climate change on water resources.

In this study, we present how the careful and consistent processing of GRACE/-FO data including a consistent estimation of the full error covariance matrix to represent the typical spatially correlated error structure supports the assimilation of satellite data into large-scale hydrological models. The first case study will tackle the question of selecting an appropriate multi-sensor data assimilation approach to combat the temporal and spatial resolution mismatch between data and model for high-dynamic frequencies. For this, daily GRACE data are assimilated for the Brahmaputra Basin that was subject to major floods, e.g., in 2004, 2007 and 2012. A reconstruction of these flood events allows a better understanding of the benefits and limitations of large-scale hydrological CDA in a changing climate. The second case study focuses on quantifying human-induced impacts on surface and groundwater storages under prolonged and intense droughts. Here, we assimilate two decades of monthly GRACE/-FO data for the Murray-Darling Basin, Australia, to better understand the impact of dry climatological conditions on our water resources.