Comparison & Integration Shell
(COINS)

An integrative modelling tool for carbon accounting and natural resource management

CRC for Greenhouse Accounting
Ecosystem Dynamics Group RSBS, ANU

Introduction | Services Provided by COINS | Key Features | Availability

Introduction

Recent advances in software and hardware technology have dramatically increased opportunities for simultaneously integrating different simulation models within the same analysis frameworks. This has many advantages, including the ability to combine different models for addressing complex environmental issues; the ability to compare model assumptions and behaviors in a consistent and standardized way; the ability to combine aspects of different models into decision support tools; and the ability to rapidly modify existing and develop new models.

This document describes the CRC for Greenhouse Accounting’s COINS model comparison and integration software. Although the software was developed specifically to aid in the analysis of terrestrial carbon dynamics, its potential uses are much more general. It is particularly suited to the analysis of the wide range of problems encountered in the broad area of natural resource management.

Background

Many mathematical models exist for investigating the dynamics of carbon in terrestrial ecosystems. They range from relatively simple ‘phenomenological’ representations of the major carbon stocks and fluxes, through to process-based models of linked carbon, nutrient and water cycles, including such detail as mathematical representations of the biophysics of photosynthesis and respiration. Despite the wide range of models and modeling approaches that have been adopted, there do exist ‘higher-level’ similarities among the various approaches. Most notable of these is the requirement to represent net biomass accumulation (or net primary productivity), and net losses via decomposition and disturbance. Another common requirement is the driver data required as model input. Most typically these comprise sets of climate data, possibly coupled with information on land cover, land-use and/or disturbance history.

Although there exist many models for investigating terrestrial carbon dynamics, comparing model behaviours and outputs is difficult because individual models are typically available only as stand-alone packages, running on a range of software platforms, each typically requiring their own specialist formatting for data input and output etc. It is therefore difficult to ensure that outputs from different models are conceptually comparable, with further difficulties in standardizing and presenting results derived from a range of software applications. Also, model development typically demands a great deal of effort in formatting datasets prior to use, in developing code for accessing that data and making it available to the ‘core’ equations of the model, and in developing other ancillary code such as file handling routines and the visualization of results. These problems are not confined to models of terrestrial carbon dynamics, but are generic across the wide range of disciplines which use simulation models as a tool for analysis.

To address the difficulties in conducting model inter-comparisons noted above, and to provide a tool within which models of terrestrial carbon dynamics can be rapidly and efficiently constructed, the CRC for Greenhouse Accounting has developed the COINS integration shell. The major aims of the COINS project were to:

1. Identify the commonalities which exist among existing models, and use this information to design software architecture which takes advantage of these shared characteristics and requirements.
2. Provide a tool to enable the consistent comparison of different models within the same modeling environment, allowing the same datasets and scenarios to be applied consistently across different models, and to allow model outputs to be presented in a consistent format.
3. Provide a research tool which can be used to investigate issues of importance in the accounting of carbon in Australian landscapes, and to provide a tool capable of further enhancing our knowledge of terrestrial carbon dynamics in the Australian context.

The COINS software is therefore not a model in itself, rather, it is one interface to many models. The software architecture within which these models reside has been optimized to allow existing and new models to be incorporated with a minimum of fuss and effort, and to provide a means of simultaneously combing many model within the same simulation environment. COINS represents the fourth in a series of software modeling shells developed primarily by Ian Davies and coworkers at the Research School of Biological Sciences, at the Australian National University. The previous products include ALEX (a meta-population modelling shell); MUSE (a vegetation modelling shell for examining the effect geometric detail has on the dynamics of vegetation models on a spatial scale of 102 - 104 m2); and LAMOS (a landscape modelling shell for exploring the interaction of landscape processes such as fire, seed dispersal and vegetation dynamics).

The main services provided by the COINS environment are summarized below, and are discussed in greater detail in the descriptions of the key features.

Services provided by the COINS environment

• The ability to incorporate models at a range of temporal and spatial scales – from days to centuries, and from ‘aspatial’ points to GIS-type simulations of continental-scale dynamics.
• The ability to simultaneously combine different models, or different parameterizations of the same model.
• Access to a range of utility tools, such as editing and manipulation of spatial and point datasets, post simulation graphical analyses of results, and access to Monte-Calro analyses for quantifying sensitivity of model outputs to uncertainty in model inputs.
• The ability to view graphically any combination of output variables, in a range of formats (e.g. maps, xy-plots and timelines).
• Rapid model development. The ability to develop models rapidly and efficiently in COINS is a result of the separation of all ancillary functions (such as data input and output routines, file handling routines) from the code specifying the core equations of the model. The model developer therefore need only be concerned with coding these core equations; all other ancillary (and time-consuming) aspects of the coding are automatically handled by the shell.
• Access to a library of established models of terrestrial carbon dynamics. The COINS model library currently includes: RothC, 3PG, 3PGs, Cen-W, Pnet2, CASS, [Century 4], Miami, RFBN, VAST, and is continually expanding.
• Access to a database of point- and continental-scale data relevant to the modeling of plant growth, litter decomposition and soil carbon dynamics. The database includes information on soil and vegetation characteristics, climate (both long-term average and monthly historical), in addition to a wide range of other data products (see http://www.greenhouse.crc.org.au/members/datasets.cfm).

Next >> Key Features