Improving evapotranspiration estimation in pasture and native vegetation models using flux tower data, remote sensing and global optimisation

J. Owens, J. Carter, G. Fraser, J. Cleverly, L. Hutley, J. Barnetson

    Research output: Chapter in Book/Report/Conference proceedingConference Paper published in Proceedings

    Abstract

    GRASP is a biophysical model of soil water balance, pasture growth and animal production developed for northern Australian grasses in wooded and non-wooded systems. The intention of this work is to improve predictions from the GRASP model of evapotranspiration, soil water balance and subsequent pasture biomass and cover in tree-grass systems. This work feeds into the operational modelling system of GRASP that is disseminated through the FORAGE and AussieGRASS online systems, available at the Long Paddock website (https://www.longpaddock.qld.gov.au/forage). The GRASP model operates at 3 different scales: Cedar GRASP (paddock scale), FORAGE (property scale) and AussieGRASS (continental scale for Australia). The Cedar version is used for model development and research on grazing trials in Queensland and the Northern Territory. FORAGE is an online system for Queensland that generates and distributes customised PDF reports with information for individual properties. Currently over 2000 reports are requested per month for use by extension providers (government and private), consultants (valuers, agents), researchers (universities and government) and land managers. AussieGRASS products are currently used within the Queensland government to assist with drought declaration assessments and a monthly Climate Outlook and Review delivered through https://www.usq.edu.au/research/environmental-sciences/qdmc-drought This paper documents the parameterisation and improvements to GRASP for estimating evapotranspiration in tree-grass systems. GRASP was overestimating the daily rate of evapotranspiration, particularly in wooded systems during the first days after rainfall events, with evapotranspiration often exceeding 1.3 times pan evaporation (Allen et al., 1998). Model partitioning of evapotranspiration into soil evaporation, grass and tree transpiration also needed adjustment to prevent excessive water loss. Incorporating daily measurements of evapotranspiration from TERN flux tower data provides the capacity to evaluate and improve the estimation of evapotranspiration in GRASP. Model changes include incorporation of satellite-derived fractional ground cover index for green and total cover in the understorey and persistent green for foliage projected cover to further improve the modelling by constraining estimates of evapotranspiration components. Combining field data with remotely sensed data and a global optimiser in an automated system provides the ability to inform model parameterisation and evaluation. Improving evapotranspiration modelling improves the soil water balance, pasture growth, tree-grass competition and safe carrying capacity, where animal numbers are matched to available pasture. Implications for these model changes and evaluation are significant, as this improves our capacity to model grazing land management issues such as runoff, export of sediment to the reef and sustainable long-term carrying capacity. Key learnings from the optimisation experiments revealed where the model needed improvements, along with careful consideration of trade-offs in regard to variable weighting when optimising multiple measured data groups (such as soil moisture, evapotranspiration and green cover). Model improvements removed the 'spikes' in daily evapotranspiration, compared well to measured data and reduced estimated tree transpiration. Daily estimates of surface soil moisture from remote sensing platforms can be used in model calibration but first require model processes and parameterisation to be appropriate at daily time steps. Calibration of evapotranspiration at a daily time step has not been tested before with GRASP due to the lack of high quality daily data sets, especially from mixed tree and grass systems. These results demonstrate the improvements in GRASP for estimating daily and monthly evapotranspiration in mixed tree and grass systems.

    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)
    Pages414-420
    Number of pages7
    ISBN (Electronic)9780975840092
    Publication statusPublished - Dec 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
    CountryAustralia
    CityCanberra
    Period1/12/196/12/19

    Fingerprint Dive into the research topics of 'Improving evapotranspiration estimation in pasture and native vegetation models using flux tower data, remote sensing and global optimisation'. Together they form a unique fingerprint.

  • Cite this

    Owens, J., Carter, J., Fraser, G., Cleverly, J., Hutley, L., & Barnetson, J. (2019). Improving evapotranspiration estimation in pasture and native vegetation models using flux tower data, remote sensing and global optimisation. In S. Elsawah (Ed.), 23rd International Congress on Modelling and Simulation - Supporting Evidence-Based Decision Making: The Role of Modelling and Simulation, MODSIM 2019 (pp. 414-420). (23rd International Congress on Modelling and Simulation - Supporting Evidence-Based Decision Making: The Role of Modelling and Simulation, MODSIM 2019). Modelling and Simulation Society of Australia and New Zealand Inc. (MSSANZ).