GFDL - Geophysical Fluid Dynamics Laboratory

Biogeochemistry, Ecosystems and Climate Group Plan


I. Introduction

II. Current Models

III. Model development and research goals over next few years

IV. Community service

I. Introduction

The main conclusions of the
Intergovernmental Panel on Climate Change (IPCC)
Fourth Assessment Report (AR4)
from Working Group 1 (WG1)

  • the warming of the planet is

  • it is very likely that the global
    temperature increase observed over the past 100 years is due to
    human activities;

  • climate change this century is
    likely to be larger than in the 20th century

These conclusions could not have been
reached without the use of climate models. Over the past several
decades, these models have progressively become more refined and
sophisticated. Current generation climate models (Atmosphere-Ocean
General Circulation Models – AOGCMs) include atmosphere, land, sea
ice and ocean components ? the main parts of the physical climate
system. These models are typically forced with estimates of past and
future changes in greenhouse gases (GHG) concentrations.

It has been long recognized that both
land and ocean biology can impact the concentration of CO2
in the atmosphere by taking up and releasing CO2.
According to the current estimates, the earth?s biosphere is
taking up about 50% of the human emissions of CO2. A very
important question is whether or not this biological uptake will
change in the future. In order to simulate the interactions of
biology and climate, earth system models (ESMs) had to be developed.

To convert an AOGCM into an ESM, one
needs to add components which can simulate the biogeochemical
processes on the land and in the ocean. The modeling of these
processes is a very new field of endeavor and, as a result, the level
of uncertainty is quite high. It is unclear what parts of the system
need to be included in the model and what parts can be neglected or
greatly simplified, and what parts are prone to bias due to the
biases in the underlying AOGCM. In many cases, there are only limited
observations of key variables needed to evaluate model performance
and to improve the models. A combination of modeling, theory and
observations are needed to make progress in earth system modeling.
Fortunately in recent years, progress is being made in all three

This gives an overview of our plans to
study the biosphere-climate interactions, starting with a description
of our current models and their development. We then present our
research and model development goals for the next several years. In
the last section, we give a description of how we see ourselves
serving the needs of both the scientific community and society.

II. Current Models

To start building the Earth System
Model (ESM)
, we used as a base a successful climate model developed
at GFDL, CM2.1. This model has been used in many climate studies
including the most recent Intergovernmental Panel on Climate Change
assessment, AR4. To this model, we added the ocean
biogeochemistry component (TOPAZ)
and replaced the land surface component with
a new land dynamics model (LM3V), which simulates vegetation dynamics
and land surface exchanges of energy, water and CO2 . This
new model is called ESM2.1.

We are using this model in several
climate studies. We have run an 1860 control integration. A control
integration is an experiment where the radiative forcing ?
Greenhouse gases, aerosols, solar, volcanoes, land use, etc. – are
all held constant. We also performed simulations of the observed
changes over the 20th century and the future changes. We
have just begun the analysis of these integrations and are preparing
papers describing results of these studies.

We are currently in the process of
improving ESM2.1 and are building two new ESMs ? ESM2M and ESM2G.
These two new models use an improved land dynamics model LM3,
with improved hydrology, snow and soil thermodynamics coupled to the
new vegetation component from LM3V. The first model, ESM2M, uses a
new version of the ocean component model (MOM4).
The second model, ESM2G, uses a different oceanic component, GOLD,
with density-based vertical coordinate.
We plan to use these models to study climate including the next IPCC

It is important to note that components
beyond traditional climate foci are included in the new ESMs such as
land use changes and the occurrence of wild fires. The need to
characterize the ecological and biogeochemical processes requires
that climate scientists at GFDL reach out to experts in many areas.
Our first collaborations have been with Princeton University
and the University of New Hampshire.

Model development and research goals over next few years

We plan to use our ESMs to study past,
present and future climates and biosphere dynamics. As noted above,
we plan to use ESM2M and ESM2G as part of GFDL?s contribution to
the IPCC AR5. We will also seek to understand past observed climate
and carbon changes. It is likely that these studies will expose
aspects of our models that need to be improved. In addition to
studying past climate changes, we will make projections of future
climates. As part of this activity, we will investigate the causes of
uncertainty in our simulations. We are also investigating the degree
to which our models can be applied to assess ecological impacts.

The model improvement and development
will continue to be an important component of our research. We are
already working on incorporating additional biogeochemical cycles
such as nitrogen into the ESMs. Nitrogen plays a crucial role in
determining plant growth rates and coastal ecosystem behavior.
Adding these new biogeochemical cycles will allow us to study new and
different facets of the impact of human activities on both the
biosphere and on climate. We are also working on improving the
capability of these models to simulate the potential climate and
anthropogenic impacts (e.g. ocean acidification and hypoxia,
terrestrial air pollution) on ecosystems through improved
representation of biodiversity and physiological and ecological

Using ESM2M and ESM2G will also allow
us to investigate the role of the ocean in both carbon and heat
uptake in the transient climate change problem. Uncertainties in
ocean heat uptake have long been known to be important for
understanding the differences among models for future projections of
climate changes. By using two different ocean component formulations,
we will be able to improve our understanding of underlying cause and
consequences of these uncertainties.

In addition, studies of the distant
past climates such as the Last Glacial Maximum (LGM ? about 21,000
years ago) will be conducted. Drawing on the experience of the LGM
study with CM2.1, we will use ESM2.1 to explore biosphere-climate
interactions and their implications for both physical climate and the
carbon cycle at that period. These activities are essential steps in
building confidence in our model projections of future climate
changes and they highlight various model strengths and weaknesses
that need to be addressed.

Community service

We anticipate our models to play an
important role in the IPCC AR5. An important part of this service is
making the model results available to the AR5 authors and the broader
climate community. We plan to make the model data available via the
data portal hosted and managed by GFDL. In addition to providing GFDL
model data, this portal will also allow users to obtain any of the
Coupled Model Intercomparison Project version 5 (CMIP5)
data sets via a network of data servers managed by DOE?s Program
for Climate Model Diagnosis and Intercomparison (PCMDI)

through the Earth System Grid (ESG).

We will also communicate our findings
through papers published in the refereed literature. This literature
will be assessed by the AR5 authors and incorporated into the IPCC
findings. In addition, we will communicate our results at various
scientific seminars, workshops and by serving on national and
international committees. GFDL scientists will also participate in
national and international assessments.

Last Updated: June 10, 2009