Land RP Report - December 2017

Status of CABLE

A key differentiation between the UK Meteorological Office model and ACCESS is the CABLE land model. A great deal of work has been invested into first implementing new science into CABLE, and then coupling this new version of CABLE to ACCESS. The new science is primarily linking a new hydrology parameterisation into CABLE, which is a blend of work by Mark Decker and Vanessa Haverd (CSIRO). The coupling work is led by Ian Harman who is in a jointly funded position between ARCCSS and CSIRO. A great deal of technical support is being provided from within CSIRO and ARCCSS.

This is complex and challenging work but it’s getting closer and the goal of having the latest version of CABLE coupled to ACCESS for CMIP-6 is still achievable. 

 

New FLUXNET tool

Anna Ukkola and Ned Haughton led the development of a community tool for processing flux tower measurements for land surface modelling. These data are indispensable for model evaluation but require significant post-processing, which is often done on an ad hoc basis by individual modelling groups around the world. The open-source R package transforms flux tower measurements into a format directly usable by LSMs and helps make these data more accessible to the wider land surface and climate modelling community.

 

New research

Arden Burrell (Centre PhD student) in collaboration with others from the centre, led new research into dryland degradation using remote sensing data. Dryland degradation is an issue of international significance as dryland regions play a substantial role in global food production. They explored a new trend detection methodology, accounts for seasonal and step changes effects that are related to significant structural changes in ecosystems, e.g. land use change (Fig. 1). The method was found to accurately capture known degradation events across a range of case studies.


Figure 1: The total change in NDVI between 1982 and 2013 given mean accumulated rainfall as determined using TSS-RESTREND

Mark Decker along with co-authors (Tangdamrongsub & Shing-Chan), submitted a paper on the use of GRACE data to improve estimation of soil moisture and groundwater in Australia. The paper uses a data assimilation framework, to combine recent advances in the representation of hydrology in CABLE (implemented by Mark), with data from the GRACE satellites. The authors demonstrate improvements in model skill, as well as the capacity to partition GRACE observations into changes in groundwater storage, surface soil moisture, & subsurface soil moisture. This framework provides valuable information for water resources and agriculture.

Following the water theme a collaboration with researchers from CSIRO Land and Water and led by Yuting Yang has used the stream flow response at the end of the Millennium drought to investigate the processes that determine how quickly stream flow returns to pre-drought levels throughout the Murray-Darling Basin. They were initially surprised to find that the dominant mechanism was not related to land surface conditions. Instead they found that the dominant control was the amount of drought-breaking rainfall. In hindsight this can be understood by the notion that it takes a lot of rainfall to end a severe drought such as the now infamous Millennium drought.

Dongqin Yin was recently involved in a collaboration with Sebastian Wolf from our centre partner ETH-Zurich with the aim of investigating the cause of the anomalous warming associated with the much publicised extreme drought in California. They focussed on trying to understand whether the observed warming was a cause of the drought or whether it was a response to the drought. The investigation involved the use of a new concept known as ‘radiative signatures’ that was pioneered by Dongqin Yin a few years earlier while she was on a PhD student exchange from China to the centre. Contrary to much media speculation the results showed that much of the warming during the Californian drought was actually a response to the drought (i.e., lack of rainfall) and was not a cause of drought per se. Their innovative use of satellite-based estimates of the surface radiative fluxes has opened up new opportunities for the use of observations to separate the cause and the effect at spatial scales of climatic relevance.

Martin De Kauwe has been involved with two new pieces of research that seek to improve our understanding of vegetation water fluxes. He led work that leveraged data from across FLUXNET to improve our understanding of the sensitivity of transpiration to stomatal conductance (Fig. 2). The outcome of this work is a benchmarking metric against which models can test assumptions about the degree of coupling between the vegetation and the atmosphere. He has also been involved in research that explored six possible mechanisms that lead to discrepancies in our understanding as we scale leaf-level water-use efficiency (WUE) to the whole ecosystem. This works adds to the growing body of research focused on WUE, the key metric that links the carbon and water cycles in land models.

 


Figure 2. Box and whisker plot (line, median; box, inter-quartile range) showing the estimated coupling coefficient (Ω) from FLUXNET data, grouped by plant functional type. Whiskers extend to 1.5 times the inter-quartile range, with dots outside of the whiskers showing outliers. Plant functional types are defined as: ENF - evergreen needle leaved forest, EBF - evergreen broadleaved forest, DBF - deciduous broadleaved forest, TRF - tropical rain forest, SAV - savanna, SHB - shrub, GRA - grasses, C3C - C3 crops, C4C - C4 crops. Values of n indicate the number of site-years for FLUXNET.

 

Sanaa Hobeichi (Centre PhD student) alongside other centre collaborators, have developed a new global gridded monthly evapotranspiration product. The Derived Optimal Linear Combination Evapotranspiration (DOLCE) product (Fig. 3) is a weighted synthesis of existing gridded evapotranspiration estimates that performs better than its constituent products when compared to ground based measurements and tested out-of-sample. The weighting approach considers both the performance of different constituent products, as well as the error covariance, or dependence, between them. DOLCE v1.0 can be downloaded from geonetwork.nci.org.au and its DOI is http://dx.doi.org/10.4225/41/58980b55b0495

 


Figure 3: Seasonal (a) global mean ET and (b) its associated uncertainty. DJF is shown in the left column and JJA in the right column.

Following up work from a few years ago, Michael Roderick has been involved in another collaboration with ETH-Zurich researchers that included a PhD student, Peter Greve, along with our long-term centre partner investigator, Sonia Seneviratne. Using a set of standard definitions of aridity they investigated the simulations from a group of climate models of changes in aridity from the last glacial maximum (CO2 ~180 ppm) to the end of the 21th century (CO2 ~900 ppm). While there were important regional differences, their main finding was that climate models generally predicted a global decline in aridity with CO2 induced warming. The paper will be out shortly in Environmental Research Letters and builds on earlier theoretical work on this important topic.

 

New papers

UNSW logo ANU logo Monash logo UMelb logo UTAS logo