CarboNA: International Studies of the North American Carbon Cycle
A Science Steering Committee has been formed consisting of carbon cycle scientists from Canada, Mexico, and the United States and government agency contacts from each country, to draft a Science Plan for CarboNA. Science questions that we will address include: 1. What's the current carbon budget of NA and adjacent oceans, including spatial structure and seasonal-to- interannual variations? 2. What mechanisms are involved? What processes control the time mean vs the interannual variability? 3. When will sinks saturate? Will they become sources? Are there surprises in store? What roles will be played by melting permafrost, boreal warming, and subtropical desertification, and tropical development? 4. What are the likely responses of terrestrial ecosystems and coastal oceans to climate change and enhanced CO2? 5. What roles will economic development, energy technology, and trade play in mitigating increases in fossil fuel emissions? In addition to the national research programs already underway in the three countries, we anticipate special collaborative projects of international scope. For example: 1. Studies of the response of terrestrial ecosystems to climate change along an ecological gradient from the Arctic to the Tropics; 2. Truly continental budgets for atmospheric greenhouse gases using data from land-based, airborne, marine, and spaceborne platforms; 3. An aggressively interdisciplinary intensive experiment to understand and quantify carbon cycle processes and budgets in the Gulf of Mexico Basin; 4. Investigation of the turrent state and likely future changes in carbon cycling in coastal ocean environments, including river inputs of POC, DOC, DIC, and nutrients; impacts on fisheries and coastal economies; exchange between coastal oceans and deep ocean basins; and air-sea gas exchange; 5. Government-level agreements on data sharing and harmonization, including but not limited to forest inventories, agricultural data, fossil fuel emissions data, land-use data, energy and population and vehicle miles traveled, flux towers, trace gas measurements, etc. CarboNA will be introduced, and the process and content of current discussions will be presented.
Progress in Obtaining Regional Estimates of the Fluxes of CO2,CH4 and N2O
In order to estimate the impact of an ecosystem on climate change, we need to better understand mass and energy exchange between the biosphere and the atmospheric layer near the surface. We will present flux measurements using the eddy covariance technique and the relaxed eddy accumulation technique of sensible heat, latent heat, CO2, CH4 and N2O for a wide range of ecosystems. A lack of energy balance closure is often observed during daytime when measuring turbulent fluxes at a single location and this lack of closure is frequently related to the mesoscale transport. Wavelet analysis, which can be used to quantify the contribution of a wide range of scales over a short distance, will be used to demonstrate the magnitude of this transport for various scalars. The possibility of using spatial averaging rather than temporal averaging, which can be used to capture some of the mesoscale transfer that cannot be measured at one point on a tower, will be examined. Using the combination of aircraft and tower-based systems we will show that even though the DeNitrification and DeComposition (DNDC) model does not predict the timing of peak N2O emissions very well, it provides fairly accurate mean annual N2O emission estimates. Finally, we will discuss the potential of using aircraft-based CH4 fluxes for validating CH4 emissions inventories at a regional scale.
A New U.S. Carbon Cycle Science Plan
The report "A U.S. carbon cycle science plan" (J. L. Sarmiento and S. C. Wofsy, U.S. Global Change Res.
Program, Washington, D. C., 1999) outlined research priorities and promoted coordinated carbon cycle
research across federal agencies in the United States for nearly a decade. Building on this framework and
subsequent reports (http://www.carboncyclescience.gov/docs.php), a working group comprised of 27 scientists
was formed in 2008 under the United States Carbon Cycle Science Program to review the 1999 Science Plan,
and to develop an updated strategy for carbon cycle research for the period from 2010 to 2020. This
comprehensive review is being conducted with wide input from the research and stakeholder communities.
The recommendations of the Carbon Cycle Science Working Group (CCSWG) will go to U.S. agency managers
who have collective responsibility for setting national carbon cycle science priorities and for sponsoring much
of the carbon cycle research in the United States. This presentation will provide an update on the ongoing
planning process, will outline the steps that the CCSWG is undertaking in building consensus towards an
updated U.S. Carbon Cycle Science Plan, and will seek input on the best ways in which to coordinate efforts
with ongoing and upcoming research in Canada and Mexico, as well as with ongoing work globally.
A model inter-comparison study of forest growth on two coastal and boreal forest landscapes in Canada
Projection of carbon stocks in Canada is presently accomplished using CBM-CFS3, an inventory-based model. We have performed a comparison exercise among 6 process-based models of forest growth (Can-IBIS, INTEC, ECOSYS, 3PG, TRIPLEX, CN-CLASS) and CBM-CFS3 as part of an effort to better capture inter-annual climate variability in the carbon accounting of Canada's forests. Comparisons were made on multi-decadal simulations for a Pacific Coastal Douglas-fir forest (2500ha, Oyster River, British Columbia) and a Boreal Black Spruce forest (3825ha, Chibougamau, Quebec). Models were initiated using reconstructions of forest composition and biomass from 1920 (Oyster River, OR) and 1928 (Chibougamau, CH), followed by transition to current forest composition as derived from recent forest inventories (OR 1999, CH 1998). Forest management events and natural disturbances over the simulation period were provided as maps and disturbance impacts on a number of carbon pools were simulated using the same transfer coefficients parameters as CBM-CFS3. Simulations were conducted from 1920 to 2006 for OR, and from 1928 to 1998 for CH. For CH, final above-ground tree biomass in 1998 was also extracted from the independent forest inventory. The coastal OR area initially contained about four times more ecosystem C than the boreal CH area. CBM- CFS3 simulations suggest a decline in ecosystem carbon by about 200 Mg C ha-1, dominated by a loss of biomass and woody debris C, over the 86-year period in OR as the entire area transitioned from coastal old- growth to second growth conditions. In CH, a smaller proportion of the area was affected by management and the CBM-CFS3 estimated a small net increase in total ecosystem C of about 11 Mg C ha-1 over 70 years, almost all attributed to increased biomass. Changes in tree biomass at CH were 10% less than estimates derived by difference between successive inventories. The source of this small simulation bias is attributable to the underlying growth and yield model, as well as to limitations of inventory methods. Overall, process- based models tracked changes in ecosystem C modeled by CBM-CFS3, but significant departures could be attributed to two possible causes. One was an apparent difficulty in reconciling the definition of various below- ground carbon pools within the different models, leading to large differences in disturbance impacts on ecosystem C. The other was in the among-model variability in the magnitude and dynamics of specific ecosystem C fluxes such as gross primary productivity and ecosystem respiration. Correlations among process-based models for weather effects on net primary productivity (NPP) were used to make recommendations about how to incorporate climate effects on growth into the CBM-CFS3 model.
Carbon in Mexican Ecosystems
The carbon (C) cycle is relevant for the understanding of the Global Climate Change. The present paper summarizes the studies conducted on terrestrial C (biomass and soil) in Mexico. The rural sector (land use and land use change, forestry, and agriculture) is the second largest contributor of CO2 emissions to the atmosphere (21%). The mean SOC status of Mexico's soils at different scales of aggregation—by climate, by pedological units, by vegetation groups, and by ecoregions—is 68.45 Mg ha-1and the SOC stored in the country soils is estimated at 13.126 × 106 Mg C, considering 1.917 × 106 km2 of land (98% of Mexico's territory). The SOC accumulated in each aggregation class depends on the combination of surface area and average SOC values. In general, the SOC values were associated with conditions that either favored biomass production (climate factors) or allowed the accumulation of SOC in the profile (soil depth). Sites were identified where the highest and lowest SOC values in the country were registered. However, this calculation requires defining the date and the land use at the exact moment the field sampling was performed. Information like this would allow evaluation of the variation of SOC over time, either with the same land use or a different one, and identification of the effect of land use change on SOC values. These studies will probably allow identification of areas with higher potential for soil carbon sequestration. As well, calculations of the potential CO2 emissions into the atmosphere due to land use change will be possible. At the present there is an effort to register carbon values simultaneously in vegetation and the soil, making feasible to learn more about the differences in the time dynamics of the two stocks—the aerial and the underground—aiming at modeling processes. Joint analysis of information generated at different spatial and time scales will be worthwhile to continue developing methodologies for modeling purposes. The contributions of the joint use of information of soil and biomass at different scales will be the introduction of the time dynamics of the phenomenon, and greater accuracy of and reduced uncertainty of the estimates.
The Role of Disturbance in the North American Terrestrial Carbon Budget: A Synthesis Activity
While disturbance plays a central role in regulating terrestrial the terrestrial carbon cycle at multiple spatial and temporal scales, its role in the North American carbon budget remains unresolved. A large number of ongoing research and monitoring programs in Canada, Mexico, and the United States are focused on developing and implementing approaches to quantify not only the spatial and temporal extent of disturbances, but the impacts of disturbance on terrestrial carbon cycling. As part of the ongoing NACP program, an effort was recently announced to carry out a synthesis of the impacts of disturbance. This paper will describe the areas being addressed by this synthesis activity, including planned outcomes.
Mapping and Aging Forest Disturbances for the United States between 1990 and 2000
Forest disturbances such as tree logging, fires and insect infestation are critical ecosystem processes, and play important roles on the carbon cycle, and forest stand age has recently been recognized as a necessary input to forest carbon cycle models. In our study, the forest disturbances in USA for the period of 1990-2000 were mapped using 400+ pairs of Landsat TM/ETM scenes, which were downloaded from the website of LEDAPS (Landsat Ecosystem Disturbance Adaptive Processing System) project, and have been re-sampled into 500 m resolution; and then the detected disturbances were separated into two five-year ages groups, with an aid of FIA (Forest Inventory and Analysis) data, which include the area of yearly new planting for each county in USA. A disturbance index (DI) was defined as the ratio of mid-infrared (MIR, band 5) to near-infrared (NIR, band 4) reflectances. The forest disturbances were identified through the Difference of Disturbance Index (DDI) between circa 2000 and 1990, where the positive DDI means disturbance and the negative DDI means regrowth. To spectrally match each pair of Landsat scenes for the purpose of accurate quantification of DDI, axis rotation performed on the plot between DI of the two matched scenes. In this way, any DDI caused by non- disturbance factors, such as seasonal change, BRDF, etc., are greatly reduced. The threshold of DDI for each TM/ETM pair is determined through the use of FIA data, following an assumption that the disturbed areas regenerate in the second year and the area of this regeneration per county approximates the regeneration statistics from FIA data for specific periods of time. Minor disturbances may be omitted due to the coarse resolution of the aggregated Landsat data, but the major stand-clearing disturbances (clear-cut harvest, fire) are captured, and the resulting continent-wide forest age map would be useful as input to regional carbon cycle models. Based on the fact that the DDI is generally higher for newer disturbed areas, the DDI values of all disturbed pixel within each county were sorted in descending order, and the two five-year ages groups, 1990-1995, and 1996-2000, were separated according to the DDI values. The spatial distribution of the detected disturbed areas was validated by MTBS fire data, which includes yearly forest fire scars from 1990-2000 in the western four States of USA (Washington, Oregon, Idaho, and California). A continent-wide forest age map is produced by combining a map from Canada and a map in USA derived from FIA data and this remote sensing-based disturbance mapping results. This continent-wide map would useful for North America carbon cycle modeling.