|
|
Focus 2:
Ecosystem Structure and Global Change
[Leader: Wolfgang
Cramer]
[Officer: TBA]
Introduction
Changes in atmospheric composition,
climate, and patterns of human land use will undoubtedly
lead to changes in both the distributions of plant and
animal species and the composition of ecosystems. These
changes in ecosystem structure will, in turn, lead to
significant changes in ecosystem physiology, such as primary
production, evapotranspiration and nutrient
cycling.
Of the driving forces of global
change, the most important for determining the distribution
and performance of organisms are the range and seasonality
of temperature, precipitation, and other environmental
factors; the intensity and frequency of severe, episodic
events, such as fires and hurricanes; and, for much of the
Earth, the group of demographic, economic, and social
pressures related to human activities. These factors,
combined with physiological responses such as sensitivity to
high CO2, longevity and ability to disperse, will
determine the future structure of the world's
ecosystems.
The goal of Focus 2 is to observe and
model this complex suite of impacts and responses so that
the pattern of change in ecosystem composition and structure
can be predicted. The work will be closely linked to the
other three Foci of GCTE and thus provides a strong
integrating function for GCTE as a whole. At the patch
scale, the integrating Activity 4 of Focus 1 will provide
the ecosystem-level physiological and biogeochemical
understanding needed for relating change in function to
change in structure. Focus 2 will build on the Focus 1
process models by extending them to longer time and larger
spatial scales, and by incorporating the other driving
forces of global change noted above.
At larger spatial scales, particularly
regional and global, the direct human driving forces of
change will come into play. Landscapes consist of mosaics of
patches where varying degrees of human activities are
superimposed on environmental heterogeneity. Specific
processes result from that spatial structure. Potential
changes in those land use patterns will in particular be
estimated through research of Focus 3 on the responses of
agricultural production systems to changing environmental
conditions. The demographic, economic, and technological
factors so critical in determining those patterns will also
be included through close links with the joint IGBP / IHDP
core project on Land-Use/Cover Change (LUCC).
The ability to predict change in
ecosystem structure and composition is being developed for
two distinct purposes:
The first, and more important,
purpose is to predict the impacts of global change on
terrestrial ecosystems in their own right (i.e.,
independent of their feedback to the atmosphere). If
human societies are to adapt to and perhaps benefit from
global change, then we must be better able to predict
what will happen to the terrestrial ecosystems on which
we depend. Thus, much of the emphasis of Focus 2 will be
on the development of a nested set of "impacts" models to
predict changes in ecosystem structure at a wide range of
scales, from patch to landscape to region. This modeling
capability will provide the basis for undertaking impacts
research in other parts of GCTE on more specific biomes
or production systems. An example is Activity 3.5, on
global change impacts on managed forests, in which
considerable modeling activity will be undertaken in
collaboration with Focus 2.
The second purpose is to further
develop Dynamic Global Vegetation Models (DGVMs) that
will capture the feedback effects of changes in ecosystem
structure and function to further atmospheric changes,
which can be linked to the general circulation models
(GCMs). GCTE aims to support the development of a
mechanistically-based dynamical model of global
vegetation for incorporation in Earth system
models.
The GCTE Synthesis Project, carried
out after six years of operational GCTE research, has shown
that the main scientific challenge of GCTE Focus 2 remains
unchanged. Despite significant advances in model development
and some success in linking those models to observations of
reality, we are currently still far away from the ability to
directly predict fundamental changes of ecosystem structure
as a consequence of global change. Significant limitations
in our understanding that have come to our awareness during
the synthesis are the following:
The connection between
physiological and structural response of vegetation to
climate and changing CO2
concentrations
The response of vegetation to
changing disturbance regimes
The aggregation of patch model
behaviour to continent-wide or global feedbacks (i.e. the
answer to the question "does the landscape scale matter
for the global response?")
Long term observations on any of
these items
Scaling up from species responses
to ecosystem level responses.
Revised Structure of Focus
2
Focus 2 is structured around three
major research activities and an integrating activity on the
GCTE contribution to the IGBP Terrestrial Transects Project.
The first of the three Activities (2.1) is based on the
development of ecosystem dynamics models at the patch scale.
This Activity links the Focus 2 effort to the other two
Foci, and provides the basis for the extension of the
predictive capability to larger spatial and temporal scales.
Activity 2.2 aims at examining processes and impacts at the
landscape scale. This involves scaling up the patch models
to landscapes and regions, primarily for predicting the
impacts of global change on the management of landscapes.
Activity 2.3 aims to develop regional and global scale
models of vegetation change for element cycles, for climate
feedback, and for regional and continental scale
comparisons.
The Terrestrial Transects Project
serves as an integrating facility and provides experimental
and observational data at a number of scales, from patch
through regional, and thus will be essential in assisting
the scaling up of ecosystem dynamics models. It also
functions as a regional "test-bed" for a range of different
models which can more consistently be compared to each other
if the underlying data sets are the same.
Figure 3 shows the structure of Focus
2. Table 1 lists the Activities and Tasks within Focus
2.
Figure 3
Activity
2.1: Patch Scale Dynamics
[Leader: James Reynolds]
The major effort of Activity 2.1 is to
predict the future composition and structure of patch scale
ecosystems under novel combinations of climate and
CO2 concentrations. These models must be based on
understanding of processes and, thus, will have strong links
to the experimental results of Focus 1. Achieving this
predictive capability requires that a suite of patch models
be developed, representing a range of complexity of
structure and function in order to simulate system processes
at different temporal and spatial scales, including
integration to landscapes and regions.
Patch models may be generally
classified as "gap" or "stand" models. In gap models,
emphasis is placed on how vegetation composition or
structure changes over time, whereas in stand models
emphasis is placed on how ecosystem function is affected by
climate forcings given a fixed species composition. Focus 2
has concentrated on the development of gap models and Focus
1 (integrating Activity 1.4) has concentrated efforts on
stand or "process" models. There are varying degrees of
overlap between these types of models with regard to how
much attention is given to functional vs. structural
components.
Functional components include
processes such as photosynthesis, acclimation, growth,
allocation, etc., whereas structural components include
species composition, canopy height and shape, plant spacing,
rooting patterns with soil depth, etc. Some models that are
rich in functional considerations virtually ignore structure
and some models that are rich in structural considerations
have limited functional aspects. In general, gap models
operate at much longer time scales and emphasize structural
features whereas stand models focus on the functional
responses of ecosystems over much shorter spatial and
temporal scales.
One of the major concerns of this
activity will be on incorporating the relative strengths of
both gap and stand models into patch scale ecosystem models.
This activity will continue previous GCTE activities as well
as initiate new activities to address key needs that have
been identified for patch level modeling:
to examine the interactive
processes between ecosystem structure and function that
will affect changes in both short and long-term dynamics
of community composition;
to develop patch models that link
descriptions of long-term changes in ecosystem structure
and function;
to continue the development and
evaluation of Plant Functional Types (PFTs) for use in
patch models;
to perform rigorous, quantitative
comparisons among different model approaches as well as
between models and long-term field data;
to establish a global network of
measurements, manipulative experiments, and modeling
studies of ecosystem dynamics along altitudinal
gradients;
to further develop the concept of
model generality via the topics of model modularity and
genericness;
to integrate multiple life forms
(e.g., grasses, shrubs, annuals) into a single,
spatially-explicit patch model;
to incorporate a
physiologically-based patch model into the Hierarchical
Patch Dynamics Model (HPDM);
to explore rules for scaling up
models from fine grained (or small extent) to coarse
grained (or large extent) based on the HPDM
paradigm.
Task 2.1.1:
Short-Term Experiments and Models of Ecosystem Structure and
Function
[Leader: Paul Leadley]
There has been a recent surge in both
experimental and modeling work on the role of structure
(i.e. biodiversity) in controlling ecosystem function. The
goal of this task is to provide a forum for modellers and
experimentalists to jointly assess the interactive processes
between ecosystem structure and function that will affect
changes in community composition.
Patch models are constructed with a
level of detail about plant and animal diversity that is
based on a priori assumptions about the importance of
biodiversity in controlling ecosystem dynamics. This is best
exemplified in gap and stand models by the amount of
information about species diversity that is considered which
ranges from many species to none (i.e., the "green smear"
approach). The need for structural detail at the species
level (i.e., diversity) has been hotly debated for many
years, but this debate has taken place largely in the
absence of quantitative analyses of the effects of diversity
on ecosystem function.
Objectives
Explicitly altering current patch
models to aggregate and disaggregate structural details
to get a better understanding of model responses to plant
and animal diversity.
Using models and experimental data
sets to test for the appropriate level of aggregation;
i.e., should models be aggregated at the level of
genotype, species, functional group or
community?
Testing models against data from
biodiversity vs. ecosystem function
experiments.
This activity will have strong linkage
to all activities in Focus 4; also Activity 1.7
(integration), LUCC and BAHC.
Proposed Timetable
1998 Workshop on Modeling
Biodiversity. Report of this workshop will serve as
background for soliciting input on goals and directions
from international community of modelers and
experimentalists.
1999 Establishment of
experimental-modeling network on
biodiversity.
1999 Submission of major
proposal(s) to fund experimental-modeling network on
biodiversity.
Task 2.1.2:
Long-Term Models of Ecosystem Structure and Function
[Leader: Harald Bugmann]
During the first phase of GCTE,
considerable progress was made with respect to the modeling
of ecosystem dynamics. However, most research on ecosystem
functioning ("stand" models) was performed in the context of
Focus 1, while ecosystem structure ("gap" models) was
treated in Focus 2. Over the past years, it has become
evident that reliable assessments of global change effects
on ecosystems require the consideration of both structural
and functional aspects within a single modeling
framework.
A major challenge for this task is to
identify the appropriate level of ecophysiological detail
that needs to be incorporated into models of long-term
ecosystem dynamics. This will be achieved by evaluating the
findings from Task 2.1.1 and also those from Focus 1,
specifically Activity 1.4, to derive scaled representations
of these detailed approaches, which then will be integrated
into gap models. Quantitative model comparisons will be an
indispensable tool in this respect.
Adding more details to the present
models tends to make parameter estimation more difficult,
and it is likely that many parameters of linked models of
ecosystem structure and function will not be known for each
species of interest. Therefore, the concept of PFTs needs to
be elaborated further, which will be done in close
collaboration with the activities of Task 2.2.1.
Since a central objective of GCTE is
to develop models that can be used for predicting ecosystem
behavior beyond current environmental conditions, model
generality is an important issue. Systematic and repeated
model comparison exercises will be launched that also
include tests of the models against independent long-term
field data.
Closely related to the issue of model
generality is model modularity and 'genericness': It will be
attempted to achieve a more generic definition of ecosystem
models, aiming particularly at the development of a
'toolbox' that can be applied across scales and
ecosystems.
Objectives
To develop patch models that link
descriptions of long-term changes in ecosystem structure
and function;
To further elaborate the concept
of Plant Functional Types (PFTs) and their incorporation
into the models mentioned above;
To perform rigorous, quantitative
comparisons among different model approaches as well as
between models and long-term field data, most notably for
sites along GCTE's Terrestrial Transects and Altitudinal
Gradients programs.
Proposed Timetable
1999 Joint workshop with Focus 1
to bring together forest gap modelers with
experimentalists and ecosystem modelers to compare and
assess the representation of physiology in gap
models.
1999 Initiation of network for
developing linked models of ecosystem structure and
function
2001 Synthesis
workshop
Task 2.1.3:
Altitudinal Gradient Studies
[Leaders: William D. Bowman and Harald
Bugmann]
Mountain regions present unique
challenges to and opportunities for global change research.
The steep slopes found in these regions give rise to some of
the sharpest environmental gradients found on Earth. Typical
characteristics of these gradients are systematic changes of
environmental parameters such as CO2, UV-B and
climate, changes of aspect and exposure, enhanced runoff and
erosion, etc. These characteristics make mountain regions
(1) particularly valuable in providing basic understanding
of ecological responses to global change; (2) susceptible to
the impacts of a rapidly changing climate, often coupled
with increasing land use pressures; and (3) likely to be
among the areas where signals of climate change impacts on
the biosphere can be detected and studied.
Mountain areas may also serve as focal
points for integrating the activities of GCTE, BAHC, and
LUCC. In a complex topography, the processes studied in each
of these programs interact closely. For example, it is
virtually impossible to consider ecological dynamics without
taking spatially explicit hydrological processes into
account, and vice versa.
This task forms the GCTE component of
the proposed IGBP Mountain Research Initiative, which is led
by BAHC and also involves START, LUCC, and PAGES (IGBP
Report No. 43). In many ways, the IGBP Mountain Research
Iniative will be complementary to the latitudinal transect
studies con-ducted under the IGBP Global Terrestrial
Transects Project. For example, while some environmental
factors change similarly along latitudinal and altitudinal
transects (e.g., temperature), others vary only along
latitudinal ones (e.g., photoperiod), and still others vary
only along altitudinal transects (e.g.,
CO2).
The proposed Mountain Work Plan is
described in detail in IGBP Report No. 43, and the related
Implementation plan will be published in 1999.
Objectives
To promote collaborative
ecological, micro-meteorological and hydrological process
studies along altitudinal gradients;
To assess and model the
interactive influence of topography and land surface
heterogeneity on the spatial patterns of soil moisture,
evapotranspiration, runoff generation and
erosion;
To establish a global network of
measurements, manipulative experiments, and modeling
studies of ecosystem dynamics along altitudinal
gradients;
To establish the requirements for
measurement techniques in harsh mountain
environments.
Proposed Timetable
1998 Workshop to develop the
related Implementation Plan (Pontresina,
Switzerland)
1999 Publication of Implementation
Plan, setup of international research network
2000/2001 Open Science Conference
of the IGBP Mountain Research Initiative
2000 Synthesis
Workshop
2002 International Year of the
Mountains, Synthesis Publications
Task 2.1.4:
Scaling Patch Models Based on the Hierarchical Patch
Dynamics Paradigm
[Leader: Jianguo Wu]
The structure, function, and dynamics
of ecological systems are determined by individual patches
and their interactions at different hierarchical levels. The
hierarchical patch dynamics paradigm (HPDP) explicitly
integrates hierarchy theory and the patch dynamics
perspective and thus has the potential to enhance our
understanding of pattern-process-scale relationships, which
will be critical to the landscape modeling activities in
Activity 2.2.
The main elements of HPDP include: (1)
ecological systems may be viewed as nested hierarchies of
patch mosaics with discrete levels; (2) dynamics of
ecological systems may be viewed as the composite dynamics
of interactive patches at different (usually adjacent)
hierarchical levels or scales; (3) pattern and processes are
reciprocally related, and both of them and their relation
are also scale dependent; (4) non-equilibrium and stochastic
processes are predominantly common in ecological systems
over different scales; and (5) while homeostatic stability
essentially does not exist in ecological systems (except
individual organisms), persistent ecological systems usually
exhibit metastability (i.e. quasi-equilibrium
states).
The HPDP has several implications for
modeling patch dynamics: (1) landscapes can be modeled as a
spatially-nested hierarchy composed of patch ecosystems.
These spatial patches may be defined based on natural
boundaries (e.g., soil types, topography, hydrologic units,
or disturbance regimes) at a particular scale (or range of
scales) where many, if not most, processes of interest
respond to. Different processes will operate at different
characteristic scales, which in turn dictate the average
size of patches relevant to them and thus give meanings to
patchiness at respective scales; (2) according to hierarchy
theory only three levels (or scales) are necessary to be
considered in a model in lieu of completeness and parsimony;
and (3) it is not necessary to assume the existence of a
stable equilibrium when one models systems that are
apparently persistent over certain time scale.
This work will provide useful
information on methods and approaches for scaling
physiological, bottom-up models to broader temporal and
spatial scales. This effort will be closely integrated with
the other Tasks in Activity 2.1 and will be coordinated with
the various tasks in Activity 2.2; thus, this work will
contribute to landscape modeling. The goals are the
following:
Objectives
to integrate multiple life forms
(e.g., grasses, shrubs, annuals) into a single,
spatially-explicit patch model;
to incorporate a
physiologically-based patch model into the hierarchical
patch dynamics model;
to examine local ecosystem
processes, recruitment and mortality, which can feed into
landscape-level models;
to explore rules for scaling up
models from fine grained (or small extent) to coarse
grained (or large extent), and to examine predictions and
applicability of hierarchical patch dynamics.
Proposed Timetable
1999 Preparatory workshop to
examine the relevance of the HPDP for modeling arid land
systems.
2000 Proposal for international
workshop on applying the HPDP in arid-lands.
Activity
2.2: Landscape Processes
[Leader: Sandra Lavorel]
Landscapes are defined as spatial
entities comprising of a set of interacting ecosystems
sharing a common broad abiotic environment (climate,
topography / land forms) and land use system. Usually, their
geographic range spans from a few to several 100
km2. In early approaches of global change, the
essential interactive nature of landscape scale processes
was ignored, and direct scaling was applied from local to
regional and global processes. However, the need to take
into account non-linear scaling, as well as the importance
of understanding and predicting global change effects on
landscapes in their own rights, have been gradually
recognized. In addition, both their scale and the broad
common characteristics (i.e., within landscape) make
landscapes amenable units for management and
planning.
Phenomena and processes that are
relevant to the landscape scale are:
Disturbance - disturbances that
propagate across the landscape (e.g. fire, herbivory,
pests and pathogens) will be highly sensitive to
landscape pattern; all disturbances contribute to the
composition and dynamics of the landscape;
Flows of energy and matter - not
all flows are likely to be a simple weighted sum over
landscape patches (as is for example the case for NPP);
some flows are expected to strongly depend on the actual
spatial arrangement of patches, barriers, etc. (e.g.
water, and other elements flowing through the
soil);
Dispersal and hence species
migration - movement of species through landscapes is
expected to depend both on their biology and on landscape
pattern, in particular habitat fragmentation and
dispersal corridors.
Changes in land use and their effects
on ecological processes can be measured in the landscape.
Climatic and atmospheric changes are also expected to have
direct and indirect effects that might be detectable at the
landscape scale. In return, landscape changes are expected
to have potential feed backs to atmospheric and possibly
climatic processes. The relative contribution of different
global change drivers and landscape scale processes are also
likely to differ between regions (e.g. Mediterranean vs.
boreal regions). This activity will have linage with BAHC,
DIS, IGAC, LUCC, PAGES
Objectives
The approach taken in this Activity is
to address the landscape issue from different angles. The
perspectives chosen are those considered as most critical to
the advancement of the general understanding and prediction
of the effects of global change on terrestrial ecosystems,
within the constraints of a three-year time frame. The four
research tasks address:
Landscape scale responses of
vegetation to changing land use and
disturbance;
Fire as a major disturbance that
will be influenced by climate, direct effects of
CO2 on vegetation, and land use, and will in
turn feed back to landscape pattern and
processes;
The interactions between species
biology and landscape patterns that will determine
migration in response to climatic change;
The effects of landscape pattern
on primary ecosystem processes (in collaboration with
Task 4.2.1).
The outcome of this Activity should be
a series of models that can be used:
to analyze interactions between
global change drivers and landscapes (theoretical
emphasis);
as tools to compare management
scenarios (applied emphasis);
as direct inputs or general rules
to be integrated into DGVMs.
Proposed Timetable
1998 Creation of Activity Web
page
1999 Joint workshop with pest
spread modellers
2000 Synthesis Workshop (venue to
be determined)
Task 2.2.1:
Responses of Vegetation to Land Use and Disturbance
[Leaders: Susan McIntyre and Sandra
Diaz]
Changing land use and climate affect
both landscape configuration and disturbance regimes. For
example, recent socio-economic changes in many semi-arid
areas of the world have been linked with increasing grazing
pressures on rangelands, and subsequent modifications in
landscape heterogeneity. Gaining a generic understanding of
the effects of changes in land use on landscape vegetation
patterns requires the development of a conceptual framework
based on the characterization of plant species response to
disturbance.
Previous GCTE research has contributed
to the development of plant functional classifications to
describe response to climate and atmospheric changes. These
classifications recognize that species can be grouped
according to similarity of response which in turn, can be
related to biological similarities. Likewise, it is proposed
that vegetation response to disturbance can be described
using a limited set of biological traits of the component
species, and that functional classifications for response to
different disturbance types can be derived. Previous
attempts at these classifications have repeatedly built
functional groups that reflect broad life forms. Therefore,
a hierarchical approach may be useful to identify relevant
traits (morphological, regeneration and disturbance
specific) within life forms (Lavorel et al. 1997). This
approach will be applied to the selection of relevant traits
used in the construction of functional classifications for
response to different disturbances.
Response to disturbance can be highly
context dependent. Vegetation response to a given change in
disturbance regime depends both on the long and short-term
disturbance history of the local flora, and the resulting
landscape patterns. For example, the local impacts of a
given grazing pressure depend on overall landscape
vegetation patterns. Therefore the identification of plant
functional groups for response to disturbance needs to
operate in a comparative manner across regions with
different disturbance and land use histories.
In developing this Task we will use
existing data, as well as conduct new experiments and field
studies. The results of the analysis of relationships
between disturbance regime and species traits will be used
to build functional groups based on species response to a
given disturbance type. These functional groups will be used
to develop landscape-scale models that analyze the
interactions between the functional group composition of
communities and landscape pattern.
Objectives
To provide a general understanding
of the response of vegetation to changes in disturbance
regime;
To set up and test a robust
methodology emphasizing repeatability and ease of
application to a wide range of regions;
To identify traits determining the
tolerance of plants to disturbance and to use these
traits to build functional classifications;
To test the generality of the
patterns identified by using regions with different
floras and different land use histories;
To use the functional
classifications to develop simulation models of landscape
dynamics under different land use scenarios.
Implementation
The implementation will be developed
around a first core network identified in the FAUST
(Functional Attributes Underlying Species Traits) core
research project, initiated at the end of 1995. FAUST
addresses the objectives of the Task through a focus on the
effects of grazing and soil disturbance on grasslands and
shrublands. The initial core network of FAUST includes
intensive studies in Mediterranean Europe (France and
Portugal),. Australia (Queensland Australia), and the United
States (Texas). These sites represent three different floras
(continents), four different climate types (arid and
sub-humid Mediterranean, arid and sub-humid tropical), acid
vs. limestone-derived basic soils, and different disturbance
regimes related to land use associated with different
evolutionary and historical disturbance backgrounds. The
expansion of FAUST to include a larger number of sites is an
immediate goal. This Task will also be expanded in order to
incorporate other ecosystems and disturbance types. One of
the priorities is to incorporate the effects of fire on
woody vegetation types.
In the second part of the
implementation (from 1998), an additional network will be
organized for the development of the modeling work. The
general objective of that group will be to analyze the
interactions between landscape pattern (for example the land
cover types present, the grain and connectivity of the
landscape pattern), vegetation initial composition, and
disturbance regimes. Specifically, simulation models will be
run for regions having different functional group spectra
and landscape patterns. The effects of changing land use and
natural disturbance regimes on different components of
landscape pattern will be analyzed. The sensitivity of the
models to the resolution in the functional groupings (many
detailed functional types vs. a few broad functional types)
will also be analyzed in the perspective of scaling up from
communities to landscapes and whole regions. This modeling
activity will be carried out in close links with the
landscape generic modeling activity undertaken in Task
2.2.4.
Proposed Timetable
1998 Task workshop aiming at a
synthesis of first results from research at different
network sites; the establishment of a cross-regional
comparison framework; and a first outline of landscape
scale model.
1999 Workshop to initiate modeling
exercise (venue to be determined), to define modeling
strategies, establish a modeling network, and to select
sites and scenarios of land use change for comparative
simulations.
1999 International Rangelands
Conference, Townsville, Australia: symposium on
"Functional classifications describing plant response to
grazing - philosophies, methods and results" and
synthesis workshop including cross-region
comparisons
1999 International Association of
Landscape Ecology (IALE) international
conference
2000 Activity synthesis
workshop
Task 2.2.2:
Relationships between Global Change and Fire Effects at
Landscape Scales
[Leader: Robert Gardner]
Because fire affects ecosystems across
a broad range of temporal and spatial scales, its inclusion
within a framework that predicts changes in vegetation
dynamics and ecosystem processes as a consequence of
climate, atmospheric and land use changes is problematic.
Fire regimes depend on the interactions between all these
drivers. It is not surprising that there are a variety of
approaches for modeling fire events and predicting the
pattern of fire intensity, severity and effects at landscape
scales. Unfortunately, no single approach appears to be
appropriate for all applications. The development of
coarse-grained fire models (e.g., time scales of years and
spatial scales of kilometers) is in its infancy. To date
comparison of the performance of coarse-grained models with
the performance of process-based fine-grained models (e.g.,
temporal scales of minutes and hours, and spatial scales of
meters to hectares), which resolve the effects of fire at
different spatial and temporal scales, have been hindered by
the lack of consistent sets of inputs for different model
formulations.
This Task assumes that "fire modeling"
is the simulation of a single or multiple fire events
without explicit consideration of community and ecosystem
processes that occur after fire(s). There are three factors
that comprise a fire event: 1) the start of a fire
(ignition), 2) the spread of that fire across the landscape
(growth), and 3) the termination of the fire event
(fire-ending event). This task will identify the inputs and
outputs of data and models (e.g., weather, succession,
ecosystem processes, etc.) required to design fire
simulation approaches that can predict fire effects over
extensive spatial and temporal scales. Along with assessment
of fire regimes produced by changing weather patterns, the
effect of human intervention (i.e., fire exclusion and
suppression policies) or fire management on fire ignition
and spread will be included in model structures. The
landscape fire model will be dynamic with respect to climate
so that changes in climate will be directly linked with
changes in the fire regime.
DGVMs operate at temporal and spatial
scales that are generally coarser than landscape models.
Model comparisons specifically designed to examine
differences due to the spatial and temporal scales of data
and simulation methods will be valuable in extrapolating
landscape models to the scales of DGVMs. Important processes
that influence fire dynamics will be identified and included
in DGVM fire module development.
Objectives
To carry out a systematic
comparison of models (or modules) predicting landscape
scale effects of fires;
To identification key processes
required to link global change with the prediction of
fire behavior and ecosystem effects across a range of
spatial and temporal scales;
To investigate the general
properties common to a broad range of ecosystem types and
recommend a general approach for simulating fire behavior
and ecosystem effects.
These objectives must be met in order
to evaluate the reliability of landscape scale models of
fire disturbance, establish a clear linkage between climate
change and fire effects for multiple ecosystem types, and
develop robust methods for inclusion of fire effects in
DGVMs.
Implementation
The implementation of this Task will
be developed around five sequential activities.
- A fire model review will be
carried out to identify key processes and information
needed to predict fire effects at a variety of spatial
and temporal scales. A detailed comparison of fire growth
models that use different spatial and temporal scales
(and therefore different simulation approaches) to start,
spread, and stop will be used to determine changes in
pattern as a consequence of the resolution of the models
and details of the processes they simulate.
- A series of workshops will be
conduced to compare the broad spectrum of ecosystem types
and modeling approaches in order to develop a general
approach for simulating fires in diverse ecosystem types.
The workshops will be designed to bring fire modellers
together with vegetation and ecosystem modellers, and
synthesize a set of standardized information that fire
models use as inputs and pass as outputs to community and
ecosystem dynamics models. In particular, these modules
will be made available for the development of generic
landscape models in Task 2.2.4.
- A comprehensive approach to fire
modeling should be applicable to changing fire regimes in
tropical systems. A test of the adequacy of current
theory and methods, and the utility of a landscape
approach developed mainly in Northern hemisphere
temperate forests, will be brought to bear on the problem
of fires in tropical systems. Data and processes unique
to tropical systems will be evaluated as input to fire
models and outputs assessed against recent patterns of
effects in tropical rainforest (e.g., Indonesia
1997).
- The sensitivity of a suite of fire
models to uncertainties in spatial and temporal input
data will be quantified by developing input parameters
and data layers with known errors. Effects on adequacy of
model predictions will be evaluated across a range of
temporal and spatial scales, as well as a range of
ecosystem types. Identification of these errors and their
effects of fire modeling output will be used to quantify
the adequacy of current data availability for assessing
fire effects.
- The results of the comparison in
steps (1)-(3) of models with a range of details and
processes, will be applied to meet coarse scale fire
modeling needs of the DGVMs. A generic fire module that
links fine-grained landscape results with the
coarser-grained predictions of DGVMs will be
developed.
Where possible in the implementation
of this Task, the IGBP transect data will be used in model
evaluation and comparison.
Proposed Timetable
1998 Development of fire model
network for communication among investigators interested
in the consequences of climate and land use change on
fire effects at landscape scales
1998 Selection of a subset of
landscape-scale models of fire for systematic comparison
of data requirements, model sensitivities, and
predictions (to be conducted as an electronic
forum)
1998 Workshop to evaluate fires in
tropical ecosystems, with special emphasis on the
adequacy of current models for assessing fire behavior
and effects in tropical systems
1999 (IALE - International
Association of Landscape Ecology - Conference in
Snowmass, USA): Workshop for initiation of ecosystem
comparison activity. Special emphasis of workshop will be
the evaluation of data uncertainties on model
predictions
1999 Workshop series to compare
fire models and assess their ability to predict fire
effects for multiple ecosystems and across broad spatial
and temporal scales
2000 Joint workshop with Activity
2.3 to linkage landscape-scale fire models (modules) with
DGVMs
2000 Final Activity workshop:
Synthesis of model comparisons
Task 2.2.3: Plant
Dispersal and Migration Modeling
[Leader: George Malanson]
To respond to millennial-scale
climatic change in the past, species shifted their ranges.
Landscape, regional, and global models need to incorporate
realistic dispersal and migration functions to capture this
process. These models are limited, however, by lack of
data-based functions of individual dispersal and
establishment. Moreover, the extent and connectivity of
habitat is being changed by humans. Species face a
double-bind: smaller, less connected habitats and a changing
climate. Fragmentation will result in the extinction of
species because reduced interactions change the long-term
dynamics, and this effect will be exacerbated by climatic
change. The interaction also creates a double-whammy for
researchers: it is difficult to differentiate the effects of
climatic change from those of fragmentation due to
disequilibrium (e.g., an "extinction debt"). Understanding
the effects of fragmentation on dispersal is critical to
assessing impacts of global change.
Successful modeling of dispersal and
migration should identify how landscape structures will
interact with climatic change and ecological processes to
determine future distributions. Current DGVMs compare
ubiquitous dispersal versus no dispersal, and current
regional and landscape dispersal models have functions based
on assumed seed shadows. Data-based functions are needed
improve landscape scale models and be scaled to DGVMs. The
sensitivity of the migration process to establishment
processes also needs to be determined.
Objectives
To compile data sets on dispersal
and recruitment patterns from which dispersal probability
distributions, or kernels, can be derived. Kernels will
describe dispersal per se; process-based establishment
functions need to be included to obtain full spatial
models of the recruitment process.
To use these data in model
comparisons on different landscapes in order to examine
how much landscape fragmentation can impede species
migration. The degree of impediment can be used to create
dispersal functional groups and identify
dispersal-limited or recruitment-limited taxa or
groups.
To examine scaling properties for
the dispersal functions on differing landscapes and to
derive rules for use in DGVMs or regional scale
vegetation models.
To tie into empirical work on
dispersal and on Quaternary migrations of
vegetation.
Implementation
To address objective (1), a core
network of empiricists and modellers will work on
synthesizing dispersal data sets. Data sets with consistent
format will be analyzed to create dispersal kernels. Needs
for further data for missing types of dispersal will be
identified. Approaches to generate data sets, such as
mechanistic aerodynamic models, or animal behavior
observations will also be investigated. Particular attention
will be devoted to seeking funds and an agency for support
of the data base.
The modeling tasks involved in
objectives (2) to (4) will be carried out by another partly
overlapping network. The approach will involve the use of
common or comparable protocols for creating heterogeneous
landscapes (random, hierarchical, or fractal patterns) on
which dispersal is simulated with shared initial kernels.
Modeling could then be extended to real landscapes with
contrasting patterns. Further comparisons will assess the
sensitivity to dispersal of a set of process- and rule-based
landscape models of vegetation dynamics. Finally, another
set of comparisons should address scaling procedures, by
analyzing the effects in theoretical models of different
renormalization rules used to aggregate cells. The results
of this theoretical approach will be used to 1) describe
specific spatial processes of interaction between landscape
pattern and dispersal kernels; 2) create dispersal
functional groups and identify dispersal-limited or
recruitment-limited taxa or groups; 3) contribute dispersal
modules to the generic landscape models developed in Task
2.2.4, and 4) develop dispersal rules for DGVMs. Current
analyses indicate that paleoecological data will provide
minimum estimates of potential migration rates. These data,
when combined with local climate/environmental constraints,
represent a valuable test of regional and landscape
models.
Throughout the development of the
Task, contacts among researchers will be maintained through
1) a small listserver to be rapidly created for forum
discussion and data exchange; 2) a web page, including an
early warning bibliographic system on work in press.
Exchanges with pest dispersal modellers will be encouraged
throughout.
Proposed Timetable
1998 Barcelona GCTE-LUCC Science
Conference: workshop to expand research and communication
network, discuss common research directions, plant-animal
interactions, landscape patterns, scaling, scenarios and
validation
1998 Working group on data
analysis and creation of dispersal kernels
1999 Workshop to synthesize
results on model comparisons
2000 Activity workshop: present
full models of migration with both climate change and
landscape heterogeneity
Task 2.2.4:
Linking Changing Landscape Pattern and Ecosystem
Function
[Leaders: Michael Apps]
Many ecological processes of interest
in global change studies, such as productivity,
biogeochemical cycling, and water and energy exchange,
operate at a number of scales. To date, many global models
on ecosystem processes are based on direct extrapolations of
process understanding at the point scale and ignore
landscape scale phenomena. This joint task deals with the
effects of the mechanisms of changes in landscape patterns
on ecosystem functioning, and how these effects are
expressed in the larger scale (landscape and higher)
functioning. This linkage is essential to improve our
understanding of, and ability to predict the impacts of
global change.
Making a model or data base spatially
interactive for upscaling requires more than giving spatial
coordinates to the landscape elements: the inclusion of
processes which couple different parts of the landscape or
that act at the scale above the individual elements is
critical. Examples of such processes include cross-scale
disturbances (e.g. fire, grazing, pests) and species
migration. Critical questions to be addressed in this Task
are:
When can indicators of ecosystem
functioning be simply aggregated as an area-weighted sum
of local values, and when, on the contrary, do
distribution and patchiness of landscape elements affect
these summations?
For a given landscape what are the
dominant mechanisms causing such non-linearity, and how
do they change over time?
The work will rest on two hypotheses:
First, for any landscape, there are a small number of
critical (keystone) cross-scale processes that cause
departures from the linear aggregation model. Second, these
processes have characteristic time and spatial scales that
are embodied in the spatial and temporal attributes of the
landscape components.
The questions can be addressed with a
series of analyses of increasing ecological complexity. At
the first level, simple aggregation, values for a process
over a landscape are calculated as the average by area of
patch values. Predictions using this linear model will be
compared with the next level of analysis, in which the size
distribution of patches is considered in addition to the
abundances. At this level, ecosystem processes causing patch
size distribution effects will need to be considered. At the
third level, the analysis will consider such scale
attributes as patch size and age distributions, as well as
the spatial organization of landscape elements (e.g.
adjacency, interspersion, and connectivity). Temporal
components may also have to be added to account for
hysteresis at the landscape scale. For example, the
neighborhood's disturbance history as well as the changes in
present environment may determine the present ecosystem
trajectory through time. The goal of this step-wise approach
is to understand and classify the conditions under which
assessments at each of those levels will suffice to explain
and predict landscape or regional processes.
Objectives
To determine the effects of
mechanisms that cause changes in landscape patterns
(complexity) on ecosystem functioning, such as
biogeochemical cycling and primary
To examine the influence of
spatially dynamic processes on regional-to-landscape
function
To predict how the interactive
effects of global change drivers on the relationship
between landscape complexity and ecosystem
functioning.
To develop heuristic models that
examine the interactions of cross-scale processes with
patch scale dynamics
Implementation
The implementation of this Task will
begin with a review of the considerable existing work on
upscaling techniques of ecosystem processes to the landscape
and regional scales. This review will identify 1) systems
that have already been shown to exhibit non-linearity vs.
additivity, 2) circumstances where an area-weighted average
is an inadequate technique for aggregation, and 3) the
spatially interactive processes that are responsible for
non-linearities.
Comparative studies will then be used
to relate aspects of landscape complexity to integrated
measures of landscape processes. The analyses will require
1) selection of techniques for classification of different
landscape patterns, 2) synoptic estimates of ecosystem
processes such as net primary productivity, gas emissions,
or movement of nutrients, and 3) identification of the
landscape key attributes (components and processes) that are
responsible for the changes in patterns and function.
Metrics including aspects of pattern (e.g., patch size,
shape, number, location) and functioning (e.g., source/sink
relationships) will be developed. These metrics will be
combined with process measurements, carried out along IGBP
transects and at intensive test sites, to determine which
and how many metrics are needed to best predict aggregated
process values at the landscape or regional scale. This Task
will make use of a highly interactive, ongoing electronic
workshop, in which data, mathematical and simulation modules
are exchanged to stimulate discussion and debate.
Improved estimations of the effects of
global change on ecosystem functioning at the landscape
scale will then be obtained by driving models of ecosystem
functioning with projections of changes in land-use,
disturbances, and migration. Such linked models should also
examine how past land-use and disturbance legacies affect
current ecosystem processes and future trajectories of land
cover change. Predicted changes in pattern and function will
be used to estimate the regional contribution to various
indicators of global functioning such as BGC feedbacks
through changes in NPP and NEP. A focus of the analysis will
be to detect dynamic responses such as time lags, threshold
effects, or the amplification of responses.
Concurrent to the development of this
main approach, a modeling sub-group will be convened to
examine the interactions of cross-scale processes with patch
scale dynamics, using a set of generic landscape models.
These models will incorporate the plant functional type
responses from task 2.2.1, disturbance models from task
2.2.2, and dispersal and migration rules from task 2.2.3.
The short-term aim of this activity is to analyze
systematically the interactions between landscape and patch
scale processes. The longer term aim will be to develop a
set of generic landscape response modules for use in
regional assessments and DGVMs.
Proposed Timetable
1998 Electronic workshop for
review of metrics of landscape complexity and processes
that introduce non-linearities at the larger scale.
Produce synthesis papers.
1998 (GCTE-LUCC Conference
Barcelona): Workshop to design specifications for a
generic landscape model
1998 Workshop to refine the
implementation plan, establish the design of modeling
studies, standardize the methodology of analyses, and
identify the initial networks of research groups and
projects.
1998 Collate data bases from case
studies and GCTE transect data sets
1999 Develop models and implement
with data from transects and test sites
2000 Activity workshop: Synthesis
of landscape functioning and generic landscape
models.
Activity
2.3: Global Vegetation Dynamics
[Leader: Wolfgang Cramer]
The overall goal of this activity is
to develop the capacity for prediction of continental and
global responses of the terrestrial biosphere to changes in
climate and land use on the time-scale of decades to
centuries. The theoretical groundwork for the modeling
framework has been established by previous GCTE activities,
and a review of the approx. four currently existing models
is part of the GCTE Synthesis Project. Key issues that have
been identified for requiring new initiatives
are:
linkage between ecophysiological
and population processes - at a level of abstraction that
allows global extrapolation,
linkage between macroclimate and
canopy processes (planetary boundary layer
fluxes),
development of a generic theory of
disturbance,
development of global data bases
that are required to calibrate, test and apply DGVMs
(soils, disturbances, climate etc.).
Task 2.3.1:
Development of Dynamic Global Vegetation Models
[Leader: Colin Prentice]
DGVMs are the state-of-the-art models
of terrestrial biosphere structure and function. They are
distinguished from models with more limited objectives, such
as biome models and terrestrial biogeochemical models, by
the fact that they explicitly couple processes on all time
scales from the physiology of net primary production up to
and including the dynamics of natural changes in vegetation
structure. DGVMs are envisaged as the primary vehicle for
the inclusion of terrestrial biosphere processes (through
both biogeophysical and biogeochemical interactions) in
comprehensive models of the Earth system. This task will
have strong links with BAHC, DIS, GAIM, IGAC, LUCC,
PAGES
Objectives
To monitor, promote and
co-ordinate the development of DGVMs (now underway in
several groups worldwide)
To foster critical comparative
testing of DGVMs using global data sets relevant to
different aspects of ecosystem structure and
function
To encourage the incorporation of
findings from more specialized modeling tasks (e.g. Tasks
2.2.2 and 2.2.3) into the framework of DGVMs
To assist the evolution of DGVMs
towards greater comprehensiveness through the inclusion
of hydrological and trace-gas components, and through
coupling to atmospheric and ocean models (in
co-ordination with Task 2.3.3)
To promote applications of DGVMs
to key Earth System problems, including palaeo and future
carbon cycle and atmospheric chemistry (in co-ordination
with Task 2.3.2 for futures)
Implementation
An informal group of DGVM developers
already exists and meets about twice a year. Currently, this
group is engaged in the first intercomparison of DGVMs.
Seven DGVMs are participating. The intercomparison focuses
on a set of three model experiments, designed around the
Hadley Centre coupled atmosphere-ocean simulation of climate
change under a business-as-usual scenario of future
greenhouse gas concentrations and sulphate aerosol
distributions. The experiments apply CO2 change
only, climate change only, and CO2 and climate
change together. The models are in all cases "spun up" to
approach equilibrium in a "pre-industrial" environment, then
forced for the entire experimental period up to 2100, and
finally run for an additional 100 years without further
change in CO2 or climate. This informal group is
currently preparing a joint article which will focus on
common patterns shown in the response of the various DGVMs
and their implications for global change and the carbon
cycle.
The DGVM Task will provide a more
formal context for this group. The group is naturally
expected to expand as more research teams become involved in
DGVM development. Nevertheless, it seems feasible to
continue meetings 1-2 times per year and to develop a joint
research agenda including collaborative publications. A more
precise plan will be worked out at the next meeting,
probably in late 1998. However, the following key features
are indicated:
There will be a new focus on model
validation. This will involve designing experiments, to
be performed by all participating DGVMs, whose aim is to
simulate some global properties of ecosystems for which
relevant data sets exist. For example, the models could
be forced by recent climatology data sets and outputs
compared with spatial patterns and interannual variations
of fPAR and runoff, with global fire data when these
become available, etc. This aspect will call for liaison
with DIS about the availability of key data sets, such as
soils data and satellite-based products.
Parametrizations of the "slow"
processes in vegetation dynamics sensu strictu ? above
all, the way in which the current community structure is
represented and feeds back on to the NPP and allocation
patterns of the constituent plants-have been developed
independently within each DGVM, are very diverse, and
have so far received little comparative scrutiny. As
DGVMs mature, it will be important to compare these
parametrizations more directly, possibly by setting up
"mini-experiments" to compare simulated natural
vegetation successions at a few locations.
Different modeling groups will
presumably have different priorities about which
extensions of DGVMs to attempt first. Nevertheless, it is
hoped there will be sufficient overlap to allow
comparative model experiments to be designed including
such disparate elements as spatially-explicit migration
(currently not included), interactive fire probabilities
(currently included only crudely or not at all),
spatially explicit hydrology, more realistic land-use
specification, and sources and sinks of the key
atmospheric trace gases: CH4, N2O,
NOx, NMHC's and CO.
During the lifetime of this Task
there will be an increasing focus on the full integration
of DGVM elements into an Earth System Modeling context.
This will call for workshops and activities to be planned
in co-operation with some quite different disciplinary
groups, possibly under the aegis of GAIM. Plans for
interfacing DGVM-based trace gas source/sink simulations
with atmospheric chemistry/transport models should be
developed in liaison with IGAC's Global Integration and
Modeling (GIM) activity. One relevant GAIM activity,
already in the planning stage, is the
GAIM/GCTE/IGAC/PAGES Palaeo Trace Gas Initiative which
plans to hold an inaugural workshop in Germany in the
last quarter of 1998: this is relevant because DGVMs will
ultimately be required in any attempt to simulate palaeo
trace gas sources, while palaeoscenarios provide
additional opportunities for testing the behaviour of
models in different climatic regimes. Other activities
more directly concerned with "futures" will presumably
also arise in co-operation with LUCC and Task 2.3.2.
Plans for such activities will be developed at the
GCTE/LUCC Science meeting "Earth's Changing Land" in
Barcelona in March 1998. Interfacing DGVMs with
atmospheric models, a major focus of Task 2.3.3 and of a
parallel, co-ordinated BAHC activity, will presumably
engage an increasing proportion of DGVM developers, and
here again future GAIM activities may crystallize around
both palaeo and future scenarios.
Task 2.3.2: Models
and Observations of Land Use Impact on Terrestrial
Vegetation at the Continental and Global Scale
[Leader: Rik Leemans]
Current developments towards DGVMs
mostly focus on physical and biological processes and
thereby ignore human land use as a driving force on
vegetation structure. The human influence on terrestrial
ecosystems is substantial and probably increasing. For
example, up to half of the land surface has been transformed
by human activities and more nitrogen is fixed by humanity
than by all natural terrestrial sources combined. Human
activities currently thus dominate many ecosystem processes
and influence all biogeochemical cycles and many feedback
processes between the terrestrial biosphere and the
atmosphere. The development of a predictive capacity for
future ecosystem development in a human-dominated world
therefore requires to incorporate land use and land-use
change comprehensively.
This task will largely be carried out
within the LUCC core project, but strong links with other
core projects are made. LUCC primarily focuses on developing
the necessary data and understanding required to understand
the interactions between major physical, ecological and
socio-economic factors that determine land use and drive
land-use change. Through the LUCC research foci and tasks
data and models will become available to provide
state-of-the-art descriptions of historic and current human
land use and the resulting land cover, as well as future
scenarios for these quantities.
Key GCTE issues that have been
identified for integrating LUCC data and models with DGVMs
are:
to provide detailed understanding
of attainable agricultural, and forestry productivity
levels together with probable future changes in order to
adequately determine global and regional land-use
potential;
to asses the role of human
activities on disturbance regimes;
to integrate land-use change
models and data sets with DGVMs.
Task 2.3.3:
Feedbacks from Broad-Scale Vegetation Change to Climate
Processes
[Leader: Jonathan Foley]
A crucial role for DGVMs is to improve
the representation of land surface processes in longer term
integrations of coupled Earth System Models. This task
monitors current developments in this field and ensures that
GCTE developments are made available for Earth System Model
developments where appropriate.
Objectives
To promote coordinated efforts
aimed at coupling state-of-the-art DGVMs with Atmospheric
General Circulation Models (AGCMs).
To promote the development of
global data sets that can be used in evaluating the
performance of coupled DGVM-AGCM models.
To foster a coordinated approach
to the evaluation and intercomparison of coupled DGVM-GCM
model simulations.
To encourage the application of
coupled DGVM-AGCM models to key Global Change issues,
including simulations of past climates, future transient
climate changes, changes in the carbon cycle, and the
role of land use and land cover change dynamics in the
Earth System.
Implementation
There are currently only a handful of
groups that are actively incorporating dynamic
representations of terrestrial ecosystems into AGCMs and
other Earth System Models. Naturally, many of these
investigators are already engaged in other GCTE activities,
including the development of more sophisticated DGVMs (Task
2.3.1). In addition, there are a few groups focused on the
development of "reduced-form" Earth System Models that will
ultimately include representations of global ecosystem
dynamics (this is the focus of BAHC Activity 7, lead by
Martin Claussen).
The first step for implementing this
GCTE Task will be to organize the existing community of
model developers. We expect to hold a first workshop in late
1999 (Madison, Wisconsin, USA) to develop an initial
strategy for coordinated model development, model evaluation
and intercomparison, and data set generation.
Specific objectives for the task will
be to address the following issues:
Integration of DGVMs and Land
Surface Process Models. Many of the existing DGVMs cannot
be directly incorporated into climate models, because
they do not included the detailed biophysical
parameterizations required to simulate short-term energy,
water, and momentum exchange with the atmosphere. It may
ultimately be necessary to incorporate the physics of the
Surface-Vegetation-Atmosphere Transfer models (SVATs,
mainly being developed by IGBP-BAHC) with DGVMs. Some of
the existing DGVMs already include SVAT
parameterizations, and could be used to illustrate the
advantages/disadvantages of the approach.
Use Coupled DGVM/AGCM Models to
Evaluate Specific Vegetation Feedback Mechanisms.
Ideally, the coupled model developers will all perform a
series of baseline simulations with their coupled
DGVM/GCM models to elucidate the role of specific
vegetation feedback mechanisms on the climate system. In
particular, this group could use their models to
investigate the effects of several hypothesized
vegetation feedback mechanisms, including: (a) changes in
albedo resulting from shifting boreal forest and tundra
boundaries, (b) changes in the extent of deserts and the
resulting changes in albedo and evapotranspiration, (c)
increases in mid-continental aridity, with the consequent
changes in vegetation cover and soil moisture, and (d)
changes in vegetation cover and evapotranspiration
directly resulting from the physiological effects of
increased CO2 concentrations.
Perform Fully Transient
Simulations with Coupled DGVM/AGCM Models. Ultimately, it
will be possible to use coupled DGVM/AGCM models to
explore fully transient aspects of climate and vegetation
interactions. An initial set of coupled models could
attempt a multi-century transient simulation of
CO2- and aerosol-induced climate change. These
simulations could be compared to benchmark transient
AGCMs climate change simulations that do not include
dynamic vegetation feedbacks.
These activities will require the
coordination of modeling activities for several years. The
exact mechanisms for implementing these activities are not
fully developed at this time. However, it is likely that a
group of model developers will organize themselves around
several of the themes discussed above. There is also the
possibility that the coupled model development group could
be operated in parallel with the DGVM development group
(Task 2.3.1), because they will have greatly overlapping
membership. A more detailed implementation plan will be
developed by an ad-hoc group of model developers in early
1998.
Integrating
Activity
IGBP Terrestrial Transects
Project
This integrating activity under Focus
2 is designed to facilitate the overall GCTE modeling
effort.
Since one of the major goals of GCTE
is to develop a set of nested ecosystem dynamics models, the
linkage of models designed for varying spatial and temporal
scales is a critical issue. For example, it is known that
aggregation errors occur in models that operate at
arbitrarily fixed spatial scales. Additional problems arise
when phenomena and processes, such as long-range migration
rates of vegetation and large-scale disturbance regimes,
that are not relevant at one temporal and spatial scale have
to be incorporated in models operating at other
scales.
The nested set of Focus 2 models will
require care in their coupling strategies and mechanisms.
Thus, it is essential that the evolving models be calibrated
and validated with real data at critical stages of their
development. The establishment of IGBP terrestrial transects
facilitates this validation procedure.
The initial transects are coincident
with the IGBP Terrestrial Transects, which are located in
the GCTE high priority biomes for biogeochemical research.
This co-location of observational/experimental sites
facilitates the linkage of physiological and ecosystem
dynamics models. Thus, in addition to the criteria which
need to be met for biogeochemical and hydrological research,
the transects should have existing patch-scale models and
data and appropriate personnel and computing
resources.
[Focus
1] [Focus
3] [Focus
4]
|
