Unifying principles for modelling PDF Print E-mail

Unifying principles for terrestrial ecosystem carbon, water and land-surface modelling


On this page you will find:

 


Unifying principles for terrestrial ecosystem carbon, water and land-surface modelling

 

Project overview

 

The distinction between terrestrial ecosystem/carbon cycle modelling and physical land-surface modelling seems increasingly artificial. We know that carbon, water and energy exchanges between ecosystems and the atmosphere are intimately coupled. Earth system models already include many different kinds of interactions between climate and the land. The coupling between photosynthesis and transpiration is at the centre of these interactions.

But this entire field is approaching a crisis of credibility. Different models are obstinately continuing to give very different results. No one seems to know why, nor do we know which results are correct. As a result, some very important questions just can’t be answered using the models we have. For example:

 

- Will reforestation and “carbon farming” decrease or increase freshwater supplies?
- How much additional CO2­ (if any!) will be put into the atmosphere by the effect of global warming on soil organic matter?
- Can more food be produced without increasing emissions of greenhouse gases?

 

Such questions are prone to “fall between the cracks” separating the disciplines of ecology, hydrology, biogeochemistry and environmental physics. This is a major barrier to progress.

To overcome it, we need better sharing of data and knowledge both within and between scientific communities. We also need a way to develop models so that it becomes possible to understand how the scientific evidence from observations and experiments informs the way the models work. Only then will it be possible to assess whether the models really are consistent with currently available data and knowledge. Generally, in one respect or another, we suspect they are not.

A lot of activity recently has gone into developing benchmark standards for land models, based on multiple observational data sets. This is essential – but it’s by far not enough. The project’s aim is to go beyond benchmarks, to set out principles which should inform the development of a new generation of models. The principles will be of two kinds:

 

- Computational principles: technical standards to promote transparency and modularity in model development, evaluation and application.
- Scientific principles: working towards a shared understanding of various key processes that overlap the disciplines.

 

The kick-off event for the project will be a workshop held under the auspices of ACEAS. Discussions at the workshop will lay out a road map for follow-up activities during the subsequent 12 months. The flagship activity will be an overview article for a scientific journal. This article will include a statement of principles, a modelling flowchart, a list of essential parameters, and a catalogue of the most relevant data sets for model evaluation and/or data assimilation. Additional activities (it is important to ensure that this project gets “traction”) will involve liaison with climate modelling centres that are planning investments in next-generation land-surface modelling: the Bureau of Meteorology (in collaboration with CSIRO) and the CABLE model in Australia, the National Center for Atmospheric Research and the CLM in the United States, and the Meteorological Office (in collaboration with the Natural Environment Research Council) and the JULES model in the United Kingdom. Another key partner in these discussions will be Microsoft Research, who are planning an ambitious global initiative in observationally-based modelling of the carbon cycle.

 

For further information about the activities of this group contact the Principal Investigator, This e-mail address is being protected from spambots. You need JavaScript enabled to view it

 

return to top


 

Workshop Report (9-13 April 2013)

The premise for this workshop was that the time has come to reformulate the principles adopted in the development, evaluation and application of models of the terrestrial biosphere. There is a range of such models – some focus mainly on the carbon dynamics of ecosystems, some on water resources, some on the ‘fast’ exchanges of energy, water and carbon between the land and atmosphere (these are the land-surface models that can be coupled to climate models), and some on the slower dynamics of vegetation change in response to a changing environment. However, all deal basically with the same set of processes, and share many common features – as well as common problems. Perhaps the most conspicuous problem is the way models can somehow all represent the present situation quite well, while diverging alarmingly when called upon to predict the future.

The core of the meeting consisted of 1.5 days of presentations (interrupted by many spirited discussions!) and 1.5 days of planning. The topics presented ranged from remote sensing of biomass trends to state-of the-art modelling of Australia’s terrestrial carbon balance. Two common themes emerged. The first was about the design of terrestrial models to include a known set of processes, and the need to make better use of the available data sets for their development, testing and improvement. The second was about the software environment for model development and, especially, the need to improve transparency and modularity. These two themes were addressed in various ways and combinations during subsequent discussions.

The key outcomes were as follows. They embody a determination to build on the remarkable degree of consensus that the meeting revealed.

 

- A journal article will be written, enumerating a series of new principles for modelling. This will be a radical document. It will propose a decisive break with past practice. It will be developed over the course of a year to allow time to generate a substantive, closely argued, authoritative statement. The writing will be led by Colin Prentice, the Principal Investigator, who plans several visits to work with other group members. In addition to statement of principles (the centrepiece), the article will include a system diagram, a Table of ‘knowns’ that should be represented in models, and a list of benchmark data sets that should constitute a minimum set of model evaluation standards.
- A strategy to communicate the Working Group’s major findings will be carried out, exploiting new initiatives under development at several climate modelling centres (including the UK Met Office and the ACCESS group comprising scientists from CSIRO and the Bureau of Meteorology). Colin Prentice and other group members will be arranging strategic meetings in the UK, USA and Australia with the aim to promote a new community approach to land-surface modelling.

 

The workshop participants greatly appreciated being housed in a beautiful tropical environment, with excellent seafood available, and a relaxed atmosphere promoting wide-ranging discussion. The ACEAS Working Group format was perfect for the purpose. The balance of specialisms, nationalities and genders was considered to be optimal. Now it remains for the group to build on the momentum we generated, and create a lasting impact.

click to enlarge
Workshop 1 Participants. Left to right: Xu Liang (University of Pittsburgh, USA), Yingping Wang (CSIRO), Colin Prentice (Macquarie University), Brad Evans (Macquarie University), Trevor Keenan (Harvard, USA), Albert van Dijk (CSIRO), Vanessa Haverd (CSIRO), Belinda Medlyn (Macquarie University), Matt Smith (Microsoft Research, Cambridge, UK), Wang Han (Macquarie University), Natalia Restrepo-Coupe (University of Technology, Sydney). Absent from photo: Ben Smith (Lund, Sweden).

 

A few invitees were not able to join the meeting but have expressed interest in participating in follow-up activities of the Working Group, namely Gab Abramowitz (University of New South Wales), Damian Barrett (CSIRO), Martyn Clark (National Center for Atmospheric Research, USA) and Sara Mikaloff Fletcher (NIWA, New Zealand).

 

Workshop 1

 

The meeting was held at Rydges Tradewinds Hotel, Cairns.

 

This e-mail address is being protected from spambots. You need JavaScript enabled to view it , Principal Investigator

 

return to top

Last Updated on Sunday, 08 February 2015 15:33