MaTepwan, 3awm©HHbiPi asTopcKi/iM npaBOM

• Atmospheric data show substantial intcrannual variability of the growth rate of CO2. This variability is not caused by variations in fossil fuel use, and thus it demonstrates dynamic behavior of the carbon-climate svstem.

• Multiple lines of evidence suggest that there exists significant interannual variability of land and ocean CO2 exchange. Terrestrial CO? exchange appears to vary with climate, responding to large annual temperature anomalies, probably dominated by the mid- and northern latitudes. Both natural net ecosystem productions (NEP) and anthropogenic biomass burning in the Tropics may vary with precipitation and the El Niño southern oscillation (ENSO) cycle.

• Intriguing evidence of trends in the seasonal cycle of CO¿, satellite "greenness" and in situ measurements of plant phenology (timing of leaf growth and death) suggests a terrestrial biotic response to climate trends and changes in the growing season.

For some time there has been energetic discussion of feedbacks to atmospheric CO? from changing ecosystem and oceanic CO¿. Many mechanisms have been advanced as dominating the responses of the oceans and ecosystems. Although trends over the next century could be large enough to trigger responses beyond those functioning over the past decade, the atmosphere tells us that ecosystems and the oceans have responded to climate during our period of observation. This provides an opportunity to evaluate which mechanisms on land (climate, carbon dioxide fertilization, nitrogen deposition, anthropogenic disturbance, etc.) may have contributed to observed changes in CO2 sources and sinks, and it offers a focus for observations over the coming years. Similarly, observations of climate, circulation, and CO2 in the oceans provide information on the effeets on CO¿ exchange of sea surfacc temperatures and circulation anomalies. The challenges of the recent observational record offer a great opportunity for Earth System modeling experiments and for both the identification and validation of key mechanisms.

7.1.2 What Will Happen to Future Terrestrial Carbon Storage?

Far i y models assumed that terrestrial carbon storage would track atmospheric CO2 via a ft factor. In this paradigm, while atmospheric CO¿ was increasing (at least up to some asymptotic level) photosynthesis would exceed respiration, and carbon storage on land would increase. Early inverse results suggested a large Northern Hemisphere sink, which suggested a more complex picture inasmuch as a COiniriven sink should have more even global distribution. Based on climate records, atmospheric CO2 data, satellite data, and models, workers such as Dai and Fung (1993) and Myneni et al. (1997) suggested that climate change and variability over the past decades could have contributed to land uptake. Wbrk on the nitrogen cycle by Bruce Peterson and Jerry Melillo (1985), Dave Schindler and Suzanne Bailey (1993), and Reth Holland et al. (1999) suggested that changes to the global nitrogen cycle could be contributing to carbon uptake in nitrogen-limited ecosystems (linking the carbon cycle to the atmospheric chemistry and transport of oxidized and reduced N compounds). The IPCC (1995)

Maiepi/tan, 3aLHwmeHHbit?i aBTopcKUM npaBOM

suggested that recovery from historical land use could be a cause of large land uptake in the northern mid-latitudes based on forest inventory data. Most recently, work using atmospheric inversion techniques by Song-miao Fan and co-workers (1998) and by Peter Rayner and others (1999) has suggested a large (even a very large) sink in regions affected by intense land use (North America, Eurasia). This has led to a much more complex piclure of controls over terrestrial carbon storage and the potential for complex Earth System interactions.

Most models of vegetation and biogeochemistry assume a procession of states ("biome" types based on a mix of plant functional types) based on a modified "optimal" response of plants' use of water and energy to capture carbon in a given climate regime. Given the central role water and carbon play in plant metabolism, this is a reasonable point of departure and may apply in the long term. However, if we consider the trajectory r of some ecosystem response variable (such as carbon storage), the response of a region to an altered environment is conditioned on (at least) three factors. These are, first, the physiological plasticity of the organisms present in the site, which can adjust rapidly (e.g., by changing leaf area). Second is the genotypic variability present in the populations, which may add significant additional flexibility in environmental responses. Third is the rate at which species and even life forms (trees vs. grasses) can change on a site. The time scales for this change can be fast, and the time scale for an optimal response is unknow n for most ecosystems, especially in the presence of intense human activity, The extent to which the response of vegetation will be optimal w ith respect to climate change is a great unknown in projections of the future carbon cycle.

In addition, the genomes of plants correspond to a large covariance matrix of plant properties. The effects of climate changes on ecosystems will depend on how a host of associated properties will change as systems arc forced by changes to water, energy, and CO?. For example, changes to grow th rates are often accompanied by changes to plant carbon chemistry and C:N ratio. We don't clearly understand how plant chemistry will change w ith phenotypic, genotypic, or successional change, and thus changes to decomposition, soil processes, and nutrient cycling are uncertain. Accompanying plant chemical changes, effects on higher trophic levels are likely. Other, more complex changes may also occur. Changes to photosynthesis and plant carbon metabolism may also be accompanied by changes to volatile organic carbon (VOC) emissions. Isoprcnc, in particular, serves to protect plant photosynthetic enzymes from high levels of light and heat, and its specific emission varies considerably between species. Thus, changes to the main carbon/water pathways could also be accompanied by changes to emissions affecting regional-global air quality. Finally, changes to the water/energy exchange characteristics of vegetation themselves feed back directly to climate via surface temperature, Bowen ratio, and albedo, and thus the trajectory of the coupled system may be a function of complex interactions.

As we consider the development of coupled models, we commit ourselves to bringing increased biological realism to ecosystem and carbon cycle modeling, and also to considering the biological covariance of changes to carbon, water, and energy processing with other aspects of ecosystem function and of ecosystem-atmosphere interaction.

Maiepwan, 3awm©HHbiPi asTopcKi/iM npaBOM

7.1.3 How Will Carbon Respond As the Oceans Change?

Consideration of ocean carbon changes to climate change aren't new in global change research. The effects of sea surface temperature on CO2 exchange have been explored in many ways. The impacts of ocean circulation changes have, likewise, received considerable attention in the context of paleo-oceanography and in view of potential climate change effects on the thermo-haline circulation, in the oceans, circulation is driven by heat and freshwater inputs and is thus tightly coupled to the atmosphere. Circulation affects carbon by the subduction of carbon-rich surface water and by the sustaining of biological productivity through the upwelling of nutrients. Changes to ocean circulation could have large effects on ocean carbon and apparently have had large effects observable in the paleorecord,

Several new questions have arisen with respect to the future of the ocean carbon cycle. First, marine biological productivity appears to be linked to the bioavailability of iron. Iron, in the surface ocean, is derived largely or mainly from Terrestrial dust. In the present day, much of this dust is derived from erosion on managed or degraded lands. Over time, dust entrainment into the atmosphere will vary with the state of climate and vegetation on land, t hus a key marine control over carbon is linked to the Earth System, that is, to climate-and-vegetation-defincd source regions in the land biosphere, and through atmospheric transport to deposition in the oceans. Thus, understanding controls over marine carbon and marine ecosystems requires a global perspective.

Most calculations of the "biological pump'" of the biological component of the marine carbon cycle rely on Redfield ratios: ratios of organic carbon to limiting nutrients. Although Redfield ratios in the ocean are astonishingly constant, they do vary. Just as in terrestrial svstems, Redfield ratios mav vary within a marine taxon as environmental mf * ^

conditions change, and they can vary as the dominant marine taxachange. These changes could occur if environmental change affects which phytoplankton taxa are best adapted to the new circumstances, possibly modulated by effects of higher trophic levels. The controls over broad biome distributions in the relatively stable environment of the land have long been a prime focus of ecological research, and a first generation of "dynamic global vegetation models" is now in use. Little similar work has been done with marine ecosystems, and global marine ecosystem models simplify biology substantially more than terrestrial models da Clearly, linking marine community ecological models to atmosphere-ocean physical models is a much-needed next step.

7.1.4 Chemistry, Biogeochemistry, and Climate

A number of the principal greenhouse gases are, unlike CO2, reactive in the atmosphere or in fact the product (ozone) of reactive species. Great progress has been made in the understanding of chemical reactions controlling the greenhouse (and "anti-greenhouse," i.e., aerosols) constituents. The sources of the principal greenhouse gases methane and nitrous oxide are approximately quantified, as are, although less exactly, the sources of many ozone precursors. We know that the atmospheric concentrations of many key atmospheric species have varied in the past, in some cases strikingly coherently with climate (as is the case for methane and N2O). This implies links between climate

Maiepwan, 3amwmeHHbifi aBTopcKiiM npaBOM


and the sources and transformations of these atmospheric species. Species related to the ozone cycle (oxidized nitrogen, some reactive organics) are also preserved in ice cores and also show both oscillations and secular changes in the past, and so the controls over atmospheric ozone precursors may well vary with climate.

This illuminates a dangerous uncertainty. Greenhouse gases other than CO? are a substantial fraction of total radiative forcing and could become more important if COi emissions are controlled. In addition, we do not understand the contemporary trends in methane and nitrous oxide quantitatively. The growth rates of these two gases have been dynamic over the historic record, and we understand neither the long-term reduction in the methane growth rate nor its rapid change in the cool year following the eruption of Mt. Pinatubo. Much of today's nitrogen trace gas emission is caused by fertilizer use in agriculture and the mobilization of nitrogen in disturbed soils undergoing land use, Human land use, in combination with climate changes, could affect f uture nitrogen trace gas emissions significantly, as well as affect the mix of direct greenhouse gases (N2O), ozone precursors (NO), and aerosol constituents (NH j). The nitrogen cycle could be a source of surprises in future radiative forcing, thus behooving us to better understand the relationship between environmental change and future trace gas emissions. A similar argument could be made for methane, whose future emissions from northern wetlands could either increase or decrease, depending on the interactions of climate change with permafrost and high-latitude hydrology.

7.2 Next Steps for Earth System Models

A first generation of models coupling physical, chemical, and biological components of the Earth System now exists, and progress is occurring at a rapid pace in incorporating new processes into models. Significant focused effort must be placed on exploring the coupled behavior of such systems. Although much work must be done to improve existing coupled model components, a parallel effort to evaluate coupled behavior involving the biogeochemical cycles is crucial. There are three obvious lines of attack. The first of these is to evaluate models against major coupled climate-biogeochemical transients in the paleorecord to ascertain whether we can reproduce the magnitude and phasing of paleo-climate and trace gas changes. Such an effort has been initiated within the IGBP as the paleo-trace gas initiative. Second, the rich record of variability in climate and biogeochemistry over the recent past should be explored. In general, this will explore the responses of the biogeochemical systems to climate forcing, rather than the effects of biogeochemistry on climate, which is key in the past and future. However, it allows the climate forcing of biogeochemistry to be both modeled and observed fairly directly. Finally, there is substantial public and scientific uncertainty about how feedbacks between radiativciy active trace gases and aerosols and climate will operate in the future. Climate affects sources and hence concentrations, and concentrations, in turn, affect climate. There is a need to begin a systematic assessment of the sensitivity of at least those biogeochemical mechanisms we understand. Such an activity has been proposed by the IGBP-Global Analysis and Modeling Program in partnership with the WCRP (the "Great Leap," denoting the ambitious first coupled calculation proposed).

Maiepnan, 3amwmeHHbiii asTopcKi/iM npaBOM

In parallel with excercising Earth System Models that embody the mechanisms we understand (however incompletely), we must also begin to evaluate how newr mechanisms may operate. Examples of such relatively new ideas include the coupling of land and ocean via iron aerosols, the constraints on terrestrial ecosystem adaptation that arise from population processes, and the potential consequences of large ecosystem changes in the oceans. Such mechanisms can be only crudely quantified in today's global models but may have influenced past dynamics and could control future processes. In addition, coupled ciimaie-biogeoehemistrv models have at best crudely captured the synergistic effects of humans via direct impacts (not through the climate system). This must be a first-order effect in land ecosystems, in w hich disturbance responses may dominate the current magnitude of land sinks and may contribute substantially to sources. It may be of grow ing significance in marine ecosystems, where human effects on higher trophic levels are today large, and in marine coastal zones, where direct human impacts may be overwhelming.

In short, there is enormous scope for scientific accomplishment in Earth System science and Earth System modeling. The challenges facing us are of great scientific difficulty, but the answers will be of great value to society.

0 0

Post a comment