Assimilating satellite soil moisture retrievals to improve operational water balance modelling

S. Tian, L. J. Renzullo*, R. Pipunic, W. Sharples, J. Lerat, C. Donnelly

*Corresponding author for this work

    Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

    Abstract

    A simple and robust method for assimilating satellite soil moisture (SM) products into the Australian Water Resources Assessment (AWRA) model was developed and tested via the community modelling system, AWRA-CMS. The method requires time series of two satellite soil moisture products, along with AWRA simulations of upper-layer soil water storage for an offline determination of weights for use in the optimal merging of models and observations via the triple collocation (TC) technique. The candidate data sources were near real-time products from the Soil Moisture Active/Passive (SMAP), Soil Moisture and Ocean Salinity (SMOS), and Advanced Scatterometer on MetOp satellite (ASCAT) production systems. Evaluation of AWRA model performance with and without data assimilation (DA) was conducted for key variables including upper-layer soil water storage, root-zone soil water storage, evapotranspiration and streamflow against in-situ networks. The comparisons demonstrated conclusively that the assimilation of satellite SM considerably improved the accuracy and representation of AWRA model surface soil moisture across Australia. The temporal correlation was increased by 0.2 correlation units on average after the assimilation compared to open-loop across in-situ SM monitoring sites. Positive impacts were found on the simulation of streamflow over majority of catchments with an increase in correlation by up to 0.4. The impact of SM assimilation on the other variables was not as significant, largely as a result of the indirect way SM assimilation imparts constraint on those variables. Finally, an investigation into the impact of SM data assimilation on forecast accuracy was conducted through driving AWRA model with forecast meteorological forcing 9 days into the future. Improved skill in estimating surface soil moisture of AWRA were found to persist up to 4 days, and likely longer. Results of this study demonstrated the benefit of constraining model outputs with satellite soil moisture observation on improving model simulation, as well as the importance of accurate initial hydrological states on improving forecast skill. Improved SM is vital for assessing and predicting water availability and assisting policy making.

    Original languageEnglish
    Title of host publication23rd International Congress on Modelling and Simulation - Supporting Evidence-Based Decision Making
    Subtitle of host publicationThe Role of Modelling and Simulation, MODSIM 2019
    EditorsS. Elsawah
    PublisherModelling and Simulation Society of Australia and New Zealand Inc (MSSANZ)
    Pages719-725
    Number of pages7
    ISBN (Electronic)9780975840092
    Publication statusPublished - 2019
    Event23rd International Congress on Modelling and Simulation - Supporting Evidence-Based Decision Making: The Role of Modelling and Simulation, MODSIM 2019 - Canberra, Australia
    Duration: 1 Dec 20196 Dec 2019

    Publication series

    Name23rd International Congress on Modelling and Simulation - Supporting Evidence-Based Decision Making: The Role of Modelling and Simulation, MODSIM 2019

    Conference

    Conference23rd International Congress on Modelling and Simulation - Supporting Evidence-Based Decision Making: The Role of Modelling and Simulation, MODSIM 2019
    Country/TerritoryAustralia
    CityCanberra
    Period1/12/196/12/19

    Fingerprint

    Dive into the research topics of 'Assimilating satellite soil moisture retrievals to improve operational water balance modelling'. Together they form a unique fingerprint.

    Cite this