Development and evaluation of 0.05° terrestrial water storage estimates using CABLE land surface model and GRACE data assimilation

Abstract

Abstract. Accurate estimation of terrestrial water storage (TWS) at a meaningful spatiotemporal resolution is important for reliable assessments of regional water resources and climate variability. Individual components of TWS include soil moisture, snow, groundwater, and canopy storage and can be estimated from the Community Atmosphere Biosphere Land Exchange (CABLE) land surface model. The spatial resolution of CABLE is currently limited to 0.5° by the resolution of soil and vegetation datasets that underlie model parameterizations, posing a challenge to using CABLE for hydrological applications at a local scale. This study aims to improve the spatial detail (from 0.5° to 0.05°) and timespan (1981–2012) of CABLE TWS estimates using rederived model parameters and high-resolution meteorological forcing. In addition, TWS observations derived from the Gravity Recovery and Climate Experiment (GRACE) satellite mission are assimilated into CABLE to improve TWS accuracy. The success of the approach is demonstrated in Australia, where multiple ground observation networks are available for validation. The evaluation process is conducted using four different case studies that employ different model spatial resolutions and include or omit GRACE data assimilation (DA). We find that the CABLE 0.05° developed here improves TWS estimates in terms of accuracy, spatial resolution, and long-term water resource assessment reliability. The inclusion of GRACE DA increases the accuracy of groundwater storage (GWS) estimates and has little impact on surface soil moisture or evapotranspiration. The use of improved model parameters and improved state estimations (via GRACE DA) together is recommended to achieve the best GWS accuracy. The workflow elaborated in this paper relies only on publicly accessible global datasets, allowing reproduction of the 0.05° TWS estimates in any study region.

Improving Runoff Simulation in the Western United States with Noah-MP and VIC

“…The two models also use different parameterizations for certain hydrological processes, including unique model equations for canopy water storage, base flow, and other processes. Both of these hydrological model structures have found extensive application both within the U.S. and internationally, as indicated by Mendoza et al (2015) and Tangdamrongsub (2023).…”

“…Here, we aim to establish parameterizations for two LSMs --the Variable Infiltration Capacity (VIC) model and the Noah-Multiparameterization (Noah-MP) LSM across the WUS. Both models have found extensive application both within the U.S. and internationally (Mendoza et al,2015;Tangdamrongsub, 2023). The approach we use involves the application of globally optimized calibration methods and regionalization, with the objective of facilitating these models to provide reliable runoff simulations.…”