Work packages

Overview and description of current (2018/2019) Work packages in TERN Landscapes

Note: This page presently includes only the Land cover related WPs. The Soils-related WPs will be added in due course.

List of Work Packages

Short title	Full title
Evapotranspiration	A process-based evapotranspiration product for Australia
Phenology	Australian landscape phenology and vegetation dynamics for climate resilience, ecosystem services, and forecasting
Rangeland condition monitoring	Reviewing the needs and capabilities for rangeland condition monitoring in Australia
GEOGLAM RAPP (global)	Operationalising the production and validation of rangelands productivity information
Seasonal fractional cover (non-DEA) and CCI Biomass	Updating Landsat fractional cover and linking TERN to international above-ground biomass initiatives
Rangeland above-ground NPP and biomass (field validation)	Establishing a national field data network for rangelands pasture productivity monitoring and modelling
Rangeland above-ground NPP dynamics (satellite and modelled products)	First look at a national dynamic rangelands pasture productivity model and looking ahead to the next version
Fire-affected areas	Fire mapping and monitoring for all Australian ecosystems
UAS instruments and platforms	Establishing a sustainable and world-leading UAS capability for TERN and Australian ecosystem research
Field data and biomass library workflows	Improving field data workflows for the past and present
Solar Induced Fluorescence (SIF) product development	First steps towards an integrated Solar Induced Fluorescence (SIF) proximal sensing capability and product development at TERN SuperSites
Bridging science and policy	Bridging science and policy, with case studies in mangroves and riparian zones
Data systems and services	Data systems and services
Coordination, reporting and management	Coordination, reporting and management

Themes

Short title	Vegetation cover (spatial/density)	Vegetation structure (height/composition)	Field data (calibration/validation)	Sensors and Instruments (R&D)	Integration and Synthesis (new, foundational activities)	Policy and Engagement	TERN cross-connections
Evapotranspiration					X
Phenology	X	X		X		X	X
Rangeland condition monitoring	X		X			X
GEOGLAM RAPP (global)	X					X
Seasonal fractional cover (non-DEA) and CCI Biomass	X	X				X	X
Rangeland above-ground NPP and biomass (field validation)			X		X	X	X
Rangeland above-ground NPP dynamics (satellite and modelled products)		X			X
Fire-affected areas	X		X			X
UAS instruments and platforms		X	X	X			X
Field data and biomass library workflows		X	X	X			X
Solar Induced Fluorescence (SIF) product development			X	X	X		X
Bridging science and policy		X				X	X

A process-based evapotranspiration product for Australia

Short title: Evapotranspiration

Principal Investigator: Tim McVicar (CSIRO)

Description: [require a paragraph or two] Satellite-based methods typically use reflectance data, which infers ET from land surface temperature differences.

Items:

1. Update to the CMRSET actual evapotranspiration (AET) production for MODIS Collection 6, including automation as new MODIS data are received.
2. Recalibrate CMRSET AET for C6 against flux tower and streamflow observations. Provide a validation report.
3. Calibrate CMRSET for MODIS NRT (should be similar to C6), Landsat, S-2 and H-8. Expect one or two can be started this FY; provide a status report.

Australian landscape phenology and vegetation dynamics for climate resilience, ecosystem services, and forecasting

Short title: Phenology

Principal Investigator: Alfredo Huete (UTS)

Description: Phenology is the study of annually recurring biological life cycle events and the drivers and controls of their periodicity. Phenology is a characteristic property of ecosystem functioning and influences local to global biogeochemical and hydrological processes including photosynthesis, water cycling, and the energy balance. Shifts in phenology depict plants’ integrated response to climate and environment, and thereby provide important landscape measures and indicators of change.

This project delivers an updated continental phenology product at 500 m, conducts validation assessments using TERN tower data, implements trial products using higher-spatial and temporal resolution satellite data, and provides a report on the community requirements for satellite-based phenology information.

Items:

1. Finalise and complete delivery of a continental phenology product at 500 m, derived from MODIS VIIRS sensor continuity from 2000-present.
2. Validate phenology timing metrics through use of TERN flux tower measures of seasonal gross primary productivity (GPP) and light use efficient (LUE).
3. Trial experimental sub-continental phenology metrics derived from Sentinel-2 (and Landsat-8) sensor at 10-30 m resolution over select areas.
4. Trial Himawari-8 to utilise daily measures of spectral greenness indices and refine phenology timing events to 1-day precision.
5. Generate a pathways to impact plan for the phenology product and metrics.

Reviewing the needs and capabilities for rangeland condition monitoring in Australia

Short title: Rangeland condition monitoring

Principal Investigator: Ken Clarke (U. Adelaide)

Description: There are a wide range of conceptions of ‘rangeland condition’ deriving from differing management perspectives. From an animal production view rangeland condition might be defined as stock carrying capacity, current biomass, ability to produce new biomass, or ability to convert rainfall into biomass. From a conservation perspective condition might be defined in more temporal and seasonal context as difference in vegetation productivity or biodiversity from expected (given current-year climatic conditions as compared to long-term records of past conditions and variability).

However, few of these aspects of rangeland condition are currently assessed at spatially meaningful scales or with repeatable quantitative methods. We propose to review the current needs for rangeland condition monitoring and management perspectives (who needs to monitor rangeland condition, and what is their goal?) and to review current methods and literature for monitoring rangeland condition. From this review we will make recommendations on appropriateness of methods for satisfying the differing management perspectives, and outline the work required to realise these methods.

Items:

1. Review and report on the current needs and methods for rangeland condition monitoring, and deliver recommendations for TERN.
2. Contribution to: Interim report on establishing a distributed UAS capability within TERN.
3. Contribution to: Test and trial UAS field practices in a practical training session with TERN Surveillance.
4. Contribution to: Initial report on effective field validation practices (grasslands and forests) for burnt area and severity mapping.

Operationalising the production and validation of rangelands productivity information

Short title: GEOGLAM RAPP (global)

Principal Investigator: Juan Guerschman (CSIRO)

Description: The GEOGLAM Rangelands and Pasture Productivity (RAPP) program applies the TERN vegetation fractional cover products to global mapping and analysis of rangelands condition. It is a working example of TERN infrastructure generating global impact. A complementary initiative is underway in Australia, supported by Dept of Agriculture and Water Resources and with partners such as the NSW Dept Primary Industries.

The RAPP implementations will benefit from better integration with TERN data sources and infrastructure, particularly for validation of spatial products using field assessments biomass and biodiversity for diverse rangeland areas. The tasks in this this WP relate to further developing the operational infrastructure to support the on-time delivery of fractional cover and related TERN products, and to develop a complementary globally-applicable NPP product for rangelands and grasslands.

Items:

1. Update the MODIS fractional cover (FC) workflow to provide operational redundancy for the products.
2. Deploy new web processing services (WPS) for efficient spatial and temporal subsetting of the FC products.
3. Demonstrate a global implementation of a GPP/NPP model for rangelands and grasslands.

Updating Landsat fractional cover and linking TERN to international above-ground biomass initiatives

Short title: Seasonal fractional cover (non-DEA) and CCI Biomass

Principal Investigator: Peter Scarth (UQ, Qld DES)

Description: The Landsat seasonal fractional cover products are one of the highest-profile remote sensing products generated by the JRSRP and TERN. They provide 25 m resolution information on the green, dry and bare ground components of Australia’s vegetation aggregated to seasons to ensure national coverage. The underlying model is validated and trained against TERN and JRSRP’s extensive site and plot network, which in turn demonstrates the value and importance of linking standardised field data collections to validated temporal/spatial products. The Landsat FC products are used by property owners, NRMs and State of Environment, and was the first TERN algorithm to be operationalised by DEA. This work package continues TERN’s support for the Landsat FC products and identifies ongoing operational and development objectives.

ESA’s Climate Change Initiative for Biomass will map the world’s above-ground biomass and associated carbon stocks predominantly in forested environments. A number of the methods employed by CCI Biomass have been tested and developed in Australia with TERN (e.g. mangrove portal and satellite-based vegetation height). This work package leverages TERN and JRSRP’s partnership with CCI Biomass to identify and inform international requirements for TERN data.

Items:

1. Contribution to the continued production of national-scale seasonal fractional cover and derivatives including persistent green.
2. Contribution to (ESA) CCI Biomass and assess updates and validation for the vegetation height product.
3. Contribution to TERN data advocacy and outreach.

Establishing a national field data network for rangelands pasture productivity monitoring and modelling

Short title: Rangeland above-ground NPP and biomass (field validation)

Principal Investigator: Luciana Porfirio (CSIRO)

Description: For any grazing enterprise, pasture production underpins everything. In Australia there are numerous methods of estimating regional-to-national and time-and-space-varying pasture productivity. These range from weather-based empirical models, to detailed physiological models and to models based on satellite imagery. The current barrier to implementing national pasture monitoring and prediction schemes is not a lack of modelling expertise but the lack field-based data, which are needed to train and evaluate models. This work package will establish a national collaboration for planning, implementing and funding a nation-wide network of pasture production monitoring sites. Initial progress towards this goal is to establish a network of interested parties and draft the scope and aims of such an activity.

Items:

1. Initial report on establishing a national field-site network for rangeland NPP and biomass.
2. Explore opportunities with TERN to collect rangeland NPP and biomass field data as per the TERN Landscapes "Effective field validation and calibration practices”.
3. Support the collation of existing and any new field data for use in calibration and validation of NPP and biomass models and satellite data products.

First look at a national dynamic rangelands pasture productivity model and looking ahead to the next version

Short title: Rangeland above-ground NPP dynamics (satellite and modelled products)

Principal Investigator: Randall Donohue (CSIRO)

Description: Establish a first national modelled rangelands dynamic NPP product. This will consolidate existing work on C-Store Pastures, a national dynamic pasture productivity modelling system, and validate this model with existing field data to provide both a ‘baseline’ product and an assessment on the adequacy of existing field data. A review will be undertaken to inform the next body of work including effective use of more field data, summary of other modelling approaches and integration or comparison with complementary satellite data products.

Items:

1. Deliver the first version of a calibrated national dynamic pasture productivity product.
2. Report on the next version of the national dynamic pasture productivity product including, for example,
   • Summary of other models and approaches.
   • Calibration and validation data requirements.
   • Integration or comparison with remote sensing optical and radar approaches.

Fire mapping and monitoring for all Australian ecosystems

Short title: Fire-affected areas

Principal Investigator: Stefan Maier (Maitec)

Description: Fire is a regular feature of Australian ecosystems. In southern Australia fires can impact land management (forests, agriculture) and urban environments, and in northern and central Australia fires can impact large areas of grazing and rangelands. In almost all Australian ecosystems fire is a major disturbance factor. Consequently, fire information is essential for understanding ecosystem processes and states. Information on fire history can be combined with other data to understand the recovery rates of different ecosystems, which can inform better management practices. Fire is also an important as a contributor to Australia’s carbon accounts.

Mapping and monitoring fire-affected regions using satellites is an efficient way to measure the extent and severity of most fire events. The daily broad-coverage provided by MODIS gives information on the approximate day-of-burning and is effective in mapping the area and severity of large fires. Higher spatial-resolution sensors such as Landsat and Sentinel-2 can capture smaller fires but their lower-temporal resolution (days to weeks) may mean they miss short-term fires, particularly when cloud is prevalent such as in northern Australia.

Ground (or airborne) mapping of fire is important for validation of all satellite-based fire products and for reporting to state and commonwealth programs. State fire agencies and volunteers regularly map fire-affected areas however they typically use different methods with, usually, unknown uncertainty. TERN is well-placed to lead an initiative on developing nationally consistent effective field validation practices that are endorsed, and can be implemented, by state fire agencies and volunteer groups.

Items:

1. Finalise the AVHRR fire frequency product, with format and metadata updates.
2. Prototype an integrated (multi-sensor) fire product that is consistent over time and space accounting for sensor-sampling biases.
3. Initial report on effective field validation practices (grasslands and forests) for burnt area and severity mapping. Consider effective practices that can be implemented by fire agencies and volunteers, to produce consistent field data suitable for remote sensing and model validation.

Establishing a sustainable and world-leading UAS capability for TERN and Australian ecosystem research

Short title: UAS instruments and platforms

Principal Investigator: Arko Lucieer (UTas)

Description: The Unmanned Aircraft System (UAS, also known as UAV or drones) sector in Australia and globally is growing rapidly. Much of the activity across industry and research uses “off the shelf” sensors for generation of RGB and multispectral orthomosaics, and 3D point clouds. The scientific literature from the past decade has demonstrated the power of UAS remote sensing for ecosystem science. However, the data collection and processing workflows are not yet sufficiently robust for long-term data collection and monitoring. This work package builds on TERN’s previous investment in defining scientifically-relevant processes for UAS-based data collection and management.

UASs contribute to calibration and validation by providing information at spatial scales relevant to community-scale ecosystem processes: providing spatial context for, or scaling-up from, field or tower-based measurements, and providing ultrahigh-resolution spatial information (<10 cm) for comparison with airborne or satellite imagery. UASs also provide new opportunities for sensor development and testing. The medium-term aim is to establish a national network of trained operators for ecosystems sciences using a set of effective practices for consistent data collection, management and product generation. This includes members of the other TERN platforms who will be trained to provide for their own UAS requirements.

Items:

1. Establish effective UAS data collection protocols and methods that can be used by field teams with limited UAS knowledge/experience.
2. Establish a data storage and metadata standard for UAS data collection and processing to standard products.
3. Test and trial UAS field practices in a practical training session with TERN Surveillance (or other TERN partners).
4. Interim report on establishing a distributed UAS capability within TERN.
5. Initiate planning and coordination for a cal/val UAS campaign with integrated field and remote sensing components.

Improving field data workflows for the past and present

Short title: Field data and biomass library workflows

Principal Investigator: Rebecca Trevithick (Qld DES)

Description: Accurate, consistent and discoverable field data are critical for calibrating and validating land cover models and remote sensing products. The Landscapes field data collections provide data, procedures and workflows developed over the course of 12 coordinated field campaigns between 2011 and 2015 (under TERN AusCover). The field data collections measure and characterise vegetation structure, composition and density in a way that is directly comparable with biophysical modelling and satellite imagery. Field data collection procedures were developed and published including the use of Open Data Kit (ODK) forms for easy-to-use, repeatable and consistent recording of data in the field. Improved automation of data collection and management has allowed the number of star-transect site records in the DES and TERN databases to increase from 1500 (pre-ODK) to more than 4000, contributed from various groups nationally.

Items:

1. Update and finalise the online resources and data sets for the previous AusCover field validation campaigns.
2. Maintain, update and develop the ODK forms and data workflows.
3. Finalise the first release of the Field Data portal.
4. Facilitate integration of the field data workflows with other TERN initiatives as appropriate.

First steps towards an integrated Solar Induced Fluorescence (SIF) proximal sensing capability and product development at TERN SuperSites

Short title: Solar Induced Fluorescence (SIF) product development

Principal Investigator: Will Woodgate (CSIRO)

Description: Solar-Induced Fluorescence (SIF) is a photosynthetic signal detectable by specialised sensors including new satellites that provides a near-instantaneous proxy for plants photosynthetic efficiency. It is especially useful for detecting a response to environmental stress such as water shortage or excess heat. This project proposes to lay the foundations to interpret the SIF signal over south eastern Australian forests, with a particular focus on scaling the leaf-level signal to the canopy by quantifying the impact of complex 3D canopy structure. A key component will be the installation of a proximal SIF measurement capability at the Tumbarumba TERN SuperSite to provide a direct link to new satellite products (NASA and ESA) and to develop algorithms to remotely monitor canopy productivity and stress.

Items:

1. Integrate a novel TERN flux tower proximal fluorescence (SIF) sensor, at Tumbarumba.
2. Design and implementation of site characterisation protocols for SIF upscaling.
3. Scaling of SIF from leaf to canopy to quantify the impact of canopy structure on SIF, including investigation of view-sun geometry effects and implications for forest-scale dynamic GPP.
4. Demonstrate links to phenology and TERN phenocam network.
5. SIF satellite product scoping study for SE Australian production forests.

Bridging science and policy

Short title: Bridging science and policy

Principal Investigators: Graciela Metternicht (UNSW), Richard Lucas (U.Aberystwyth/UNSW)

Description: Demonstrate how excellence of science produced by TERN Landscapes can aid evidence-based policy. The 2016 Australia State of the Environment Report ascertained that “Significant quantities of environmental data are collected by individual managers, community groups and private companies. So far, the data are rarely in information systems that are accessible to others, but there have been significant investments in building the infrastructure and capacity to facilitate sharing of data”. The latter is true of TERN and its precedent efforts in building cross-institutional, open-access data and infrastructure sharing. Using mangroves as a demonstrator, the project will investigate how an Australian ecosystem monitoring and management system might be constructed and implemented given that environmental issues span jurisdictional domains. It is expected to inform TERN on approaches to monitoring other ecosystems throughout Australia.

Items:

1. Engagement at the Federal, State and Territory Level to establish a mechanism to encourage uptake, adoption and recognition of TERN Mangroves.
2. Mangroves as demonstrator for the Australian Plant Biomass Library.
3. Proof of concept for a framework to monitor dynamics of wetland and riparian zones, building on the experience of TERN Mangroves.
4. Policy implications of change:
   • How remote sensing and landscape information can help to review successes of policies.
   • Develop DPSIR (drivers, pressures, state, impact, response) extensions for the EODESM software and apply to TERN Mangroves.
5. Support TERN Landscape senior management in strategic planning and advocacy.

Data systems and services

Short title:' Data systems and services

Principal Investigators: Matt Paget, Peter Wilson (CSIRO)

Description: The TERN Landscapes data systems and services team is the interface between Landscapes partners and TERN Data Services platform. The team will facilitate the transition of services to be managed by TERN Data Services, and support and liaise with Landscapes partners to connect their data workflows to TERN-managed services. The team will also identify and support development of bespoke or custom services that are particular to the Landscapes platform.

Items:

1. Develop data workflow diagrams per data product; update data and metadata for each product.
2. Facilitate the transition of selected Soils and AusCover services to TERN Data Services.
3. Update configuration and automate management of Landscapes systems and services (those not transitioned to TERN Data Services).

Coordination, reporting and management

Short title:' Coordination, reporting and management

Principal Investigators: Matt Stenson (CSIRO), Rowena Smith (ANU), Ross Searle (CSIRO), Matt Paget (CSIRO)

Description: Project management and governance.

Items:

1. Execution, reporting and delivery of sub-contracts.
2. Reporting to the TERN Central office and Board.
3. Support TERN Central communications, website updates, activities and promotion of TERN Landscapes.
4. Management and coordination of a TERN Landscapes annual science symposium.
5. Promotion of TERN data with ANU Summer Institute and associated partnerships.
6. New version and updates to the TERN Landscapes Cal/Val field handbook.