Modeling and Spatially Distributing Forest Net Primary Production at the Regional Scale

Abstract

Forest, agricultural, rangeland, wetland, and urban landscapes have different rates of carbon sequestration and total carbon sequestration potential under alternative management options. Changes in the proportion and spatial distribution of land use could enhance or degrade that area’s ability to sequester carbon in terrestrial ecosystems. As the ecosystems within a landscape change due to natural or anthropogenic processes, they may go from being a carbon sink to a carbon source or vice versa. Satellite image analysis has been tested for timely and accurate measurement of spatially explicit land use change and is well suited for use in inventory and monitoring of terrestrial carbon. The coupling of Landsat Thematic Mapper (TM) data with a physiologically based forest productivity model (PnET-II) and historic climatic data provides an opportunity to enhance field plot-based forest inventory and monitoring methodologies. We use periodic forest inventory data from the U.S. Department of Agriculture (USDA) Forest Service’s Forest Inventory and Analysis (FIA) Program to obtain estimates of forest area and type and to generate estimates of carbon storage for evergreen, deciduous, and mixed-forest classes. The area information is used in an accuracy assessment of remotely sensed forest cover at the regional scale. The map display of modeled net primary production (NPP) shows a range of forest carbon storage potentials and their spatial relationship to other landscape features across the southern United States. This methodology addresses the potential for measuring and projecting forest carbon sequestration in the terrestrial biosphere of the southern United States.     

Regional estimation of current and future forest biomass

The 90,674 wildland fires that burned 2.9 million ha at an estimated suppression cost of $1.6 billion in the United States during the 2000 fire season demonstrated that forest fuel loading has become a hazard to life, property, and ecosystem health as a result of past fire exclusion policies and practices. The fire regime at any given location in these regions is a result of complex interactions between forest biomass, topography, ignitions, and weather. Forest structure and biomass are important aspects in determining current and future fire regimes. Efforts to quantify live and dead forest biomass at the local to regional scale has been hindered by the uncertainty surrounding the measurement and modeling of forest ecosystem processes and fluxes. The interaction of elevated CO2 with climate, soil nutrients, and other forest management factors that affect forest growth and fuel loading will play a major role in determining future forest stand growth and the distribution of species across the southern United States. The use of satellite image analysis has been tested for timely and accurate measurement of spatially explicit land use change and is well suited for use in inventory and monitoring of forest carbon. The incorporation of Landsat Thematic Mapper data coupled with a physiologically based productivity model (PnET), soil water holding capacity, and historic and projected climatic data provides an opportunity to enhance field plot based forest inventory and monitoring methodologies. We use periodic forest inventory data from the USDA Forest Service's Forest Inventory and Analysis (FIA) project to obtain estimates of forest area and type to generate estimates of carbon storage for evergreen, deciduous, and mixed forest classes for use in an assessment of remotely sensed forest cover at the regional scale for the southern United States. The displays of net primary productivity (NPP) generated from the PnET model show areas of high and low forest carbon storage potential and their spatial relationship to other landscape features for the southern United States. At the regional scale, predicted annual NPP in 1992 ranged from 836 to 2181 g/m2/year for evergreen forests and 769-2634 g/m2/year for deciduous forests with a regional mean for all forest land of 1448 g/m2/year. Prediction of annual NPP in 2050 ranged from 913 to 2076 g/m2/year for evergreen forest types to 1214-2376 g/m2/year for deciduous forest types with a regional mean for all forest land of 1659 g/m2/year. The changes in forest productivity from 1992 to 2050 are shown to display potential areas of increased or decreased forest biomass. This methodology addresses the need for spatially quantifying forest carbon in the terrestrial biosphere to assess forest productivity and wildland fire fuels.     

Modeling and spatially distributing forest net primary production at the regional scale

Forest, agricultural, rangeland, wetland, and urban landscapes have different rates of carbon sequestration and total carbon sequestration potential under alternative management options. Changes in the proportion and spatial distribution of land use could enhance or degrade that area's ability to sequester carbon in terrestrial ecosystems. As the ecosystems within a landscape change due to natural or anthropogenic processes, they may go from being a carbon sink to a carbon source or vice versa. Satellite image analysis has been tested for timely and accurate measurement of spatially explicit land use change and is well suited for use in inventory and monitoring of terrestrial carbon. The coupling of Landsat Thematic Mapper (TM) data with a physiologically based forest productivity model (PnET-II) and historic climatic data provides an opportunity to enhance field plot-based forest inventory and monitoring methodologies. We use periodic forest inventory data from the U.S. Department of Agriculture (USDA) Forest Service's Forest Inventory and Analysis (FIA) Program to obtain estimates of forest area and type and to generate estimates of carbon storage for evergreen, deciduous, and mixed-forest classes. The area information is used in an accuracy assessment of remotely sensed forest cover at the regional scale. The map display of modeled net primary production (NPP) shows a range of forest carbon storage potentials and their spatial relationship to other landscape features across the southern United States. This methodology addresses the potential for measuring and projecting forest carbon sequestration in the terrestrial biosphere of the southern United States.     

Monitoring land-use change effects on soil carbon in New Zealand: quantifying baseline soil carbon stocks

We designed a soil carbon monitoring system for New Zealand using country-specific land use and soil carbon information. The system pre-stratifies the country by soil type, climate, and land use. Soils were placed in six IPCC soil categories; Podzols were added as they are widespread throughout New Zealand. Temperature was stratified into two categories, each spanning 7 degrees C. Moisture categories were based on water balance, and included five categories. Temperature and moisture stratification was based on the USDA Soil Classification system. Land use (10 categories) was based on 1980s survey data. Overall, 39 combinations of these three factors (cells) described 93% of the New Zealand landscape. Geo-referenced soil carbon data (carbon concentration and bulk density) were used to quantify average soil carbon for each of the 39 cells. Aggregating the polygons gave an estimated 1990 soil carbon baseline of 1152+/-44, 1439+/-73, and 1602+/-167 Mt C (mean+/-S.D.) for the 0-0.1, 0.1-0.3, and 0.3-1.0 m depth increments (not including forest floor carbon). The system described could also be used to quantify equilibrium changes in soil C associated with land-use change if land use is updated periodically.     

Comparison of regression coefficient and GIS-based methodologies for regional estimates of forest soil carbon stocks

Estimates of forest soil organic carbon (SOC) have applications in carbon science, soil quality studies, carbon sequestration technologies, and carbon trading. Forest SOC has been modeled using a regression coefficient methodology that applies mean SOC densities (mass/area) to broad forest regions. A higher resolution model is based on an approach that employs a geographic information system (GIS) with soil databases and satellite-derived landcover images. Despite this advancement, the regression approach remains the basis of current state and federal level greenhouse gas inventories. Both approaches are analyzed in detail for Wisconsin forest soils from 1983 to 2001, applying rigorous error-fixing algorithms to soil databases. Resulting SOC stock estimates are 20% larger when determined using the GIS method rather than the regression approach. Average annual rates of increase in SOC stocks are 3.6 and 1.0 million metric tons of carbon per year for the GIS and regression approaches respectively.     

Carbon storage in northeast China as estimated from vegetation and soil inventories

We have estimated the stocks of carbon in vegetation and soil in northeast China based on data for 122 plots from the fourth national forest inventory, and for 388 soil profiles from the second national soil survey. The techniques of Geographic Information System (GIS) have been used to extrapolate site-specific estimates of vegetation and soil organic carbon to the entire area of northeast China. Our estimate indicates that the amount of carbon in vegetation and soil for the region are 2.81 PgC (10(15) g C) and 26.43 PgC, respectively, and that the area weighted average density of vegetation and soil organic carbon are 22.7 MgC/ha and 212.7 MgC/ha, respectively. The eastern and northern parts of the region show much higher carbon storage than the rest of the region. Substantial spatial variations in vegetation and soil organic carbon across northeast China suggest that regional estimates on carbon stocks and fluxes should take into account these spatial variations. We suggest that the methodology developed can be used for the entire nation of China as well as other regions of the world.     

Estimating down deadwood from FIA forest inventory variables in Maine

Down deadwood (DDW) is a carbon component important in the function and structure of forest ecosystems, but estimating DDW is problematic because these data are not widely available in forest inventory databases. However, DDW data were collected on USDA Forest Service Forest Inventory and Analysis (FIA) plots during Maine's 1995 inventory. This study examines ways to predict DDW biomass from other FIA variables so that DDW could be estimated without tedious measurement. Our results include a regression model that predicts DDW as a function of stand size class, basal area of dead and cut trees, and dummy variables for forest type and forest industry ownership. We also found DDW similar to FIA's standing-tree mortality at a statewide scale.     

Net carbon flux from agricultural ecosystems: methodology for full carbon cycle analyses

Agricultural ecosystems have the potential to sequester carbon in soils by altering agricultural management practices (i.e. tillage practice, cover crops, and crop rotation) and using agricultural inputs (i.e. fertilizers and irrigation) more efficiently. Changes in agricultural practices can also cause changes in CO2 emissions associated with these practices. In order to account for changes in net CO2 emissions, and thereby estimate the overall impact of carbon sequestration initiatives on the atmospheric CO2 pool, we use a methodology for full carbon cycle analysis of agricultural ecosystems. The analysis accounts for changes in carbon sequestration and emission rates with time, and results in values representing a change in net carbon flux. Comparison among values of net carbon flux for two or more systems, using the initial system as a baseline value, results in a value for relative net carbon flux. Some results from using the full carbon cycle methodology, along with US national average values for agricultural inputs, indicate that the net carbon flux averaged over all crops following conversion from conventional tillage to no-till is -189 kg C ha(-1) year(-1) (a negative value indicates net transfer of carbon from the atmosphere). The relative net carbon flux, using conventional tillage as the baseline, is -371 kg C ha(-1) year(-1), which represents the total atmospheric CO2 reduction caused by changing tillage practices. The methodology used here illustrates the importance of (1) delineating system boundaries, (2) including CO2 emissions associated with sequestration initiatives in the accounting process, and (3) comparing the new management practices associated with sequestration initiatives with the original management practices to obtain the true impact of sequestration projects on the atmospheric CO2 pool.     

Using forest health monitoring data to integrate above and below ground carbon information

The national Forest Health Monitoring (FHM) program conducted a remeasurement study in 1999 to evaluate the usefulness and feasibility of collecting data needed for investigating carbon budgets in forests. This study indicated that FHM data are adequate for detecting a 20% change over 10 years (2% change per year) in percent total carbon and carbon content (MgC/ha) when sampling by horizon, with greater than 80% probability that a change in carbon content will be determined when a change has truly occurred (P < or = 0.33). The data were also useful in producing estimates of forest floor and soil carbon stocks by depth that were somewhat lower than literature values used for comparison. The scale at which the data were collected lends itself to producing standing stock estimates needed for carbon budget development and carbon cycle modeling. The availability of site-specific forest mensuration data enables the exploration of above ground and below ground linkages.     

Soil carbon protection in podocarp/hardwood forest, and effects of conversion to pasture and exotic pine forest

A combination of paired site, time series, and survey approaches were used to estimate the effect of land use change on mineral soil carbon (C), and to identify factors associated with variation. Land-uses compared included podocarp/hardwood forest, improved pasture, and pine plantation. Soil C was significantly related to soil pH that ranged between 3.9-5.9 (0-0.05 m), 3.6-6.0 (0.05-0.10 m), and 4.5-6.1 (0.10-0.50 m) in indigenous forest. Time series data obtained by periodically re-sampling soil (0-0.10 m) in permanent plots in a pine forest previously under pasture showed that mineral soil C decrease by approximately 4 Mg ha(-1) by the end of the first rotation. The time series data compared closely with mean results obtained at paired-site throughout New Zealand. Soil C concentration was highly variable in all land-uses, and the evidence suggests that chemical stabilisation of C occurred under acid conditions in native forest, through complexation with Al, and that effects persisted long after conversion of the native forest to other land-uses. The implications of these findings for the design of sampling protocols for soil C are discussed.     

Sensitivity of carbon sequestration costs to soil carbon rates

Modifying current agricultural management practices as a means of sequestering carbon has been shown to be a relatively low cost way to offset greenhouse gas emissions. In this paper we examine the sensitivity of the estimates of the amount of soil carbon sequestered and the implied costs of sequestering a tonne of carbon to changes in the rates of soil carbon sequestered for alternative production practices. An application is made to the dryland grain production systems of the US Northern Plains where the marginal costs of soil C range from $20 to $100 per MT. We show that the resulting changes in the marginal costs quantities of C sequestered are not a monotonic transformation of the changes in the soil carbon rates. These results underscore the importance of using a linked economic and biophysical simulation model to assess the economic potential for sequestering carbon in agricultural soils.     

Soil carbon pools and fluxes in urban ecosystems

The transformation of landscapes from non-urban to urban land use has the potential to greatly modify soil carbon (C) pools and fluxes. For urban ecosystems, very little data exists to assess whether urbanization leads to an increase or decrease in soil C pools. We analyzed three data sets to assess the potential for urbanization to affect soil organic C. These included surface (0-10 cm) soil C data from unmanaged forests along an urban-rural gradient, data from "made" soils (1 m depth) from five different cities, and surface (0-15 cm) soil data of several land-use types in the city of Baltimore. Along the urban-rural land-use gradient, we found that soil organic matter concentration in the surface 10 cm varied significantly (P=0.001). In an analysis of variance, the urban forest stands had significantly (P=0.02) higher organic C densities (kg m(-2) to 1 m depth) than the suburban and rural stands. Our analysis of pedon data from five cities showed that the highest soil organic C densities occurred in loamy fill (28.5 kg m(-2)) with the lowest occurring in clean fill and old dredge materials (1.4 and 6.9 kg m(-2), respectively). Soil organic C densities for residential areas (15.5 +/- 1.2 kg m(-2)) were consistent across cities. A comparison of land-use types showed that low density residential and institutional land-uses had 44 and 38% higher organic C densities than the commercial land-use type, respectively. Our analysis shows that as adjacent land-use becomes more urbanized, forest soil C pools can be affected even in stands not directly disturbed by urban land development. Data from several "made" soils suggests that physical disturbances and inputs of various materials by humans can greatly alter the amount C stored in these soils.     

A carbon balance method for paper and wood products

The approach used to track the flow of carbon sequestered in the forest through harvest, processing into products, and final disposition of products is described. The methodology is broadly flexible and applicable to forest-based carbon balance assessments. A carbon balance is computed across all forestland ownerships for the production facility of interest. The balance considers forest uptake, harvest, combustion of fuels, emissions from process steps and losses from product use, disposition and recycling. The method also allows for sensitivity and marginal assessments of a variety of real and hypothetical situations using variable assumptions. Example results for a vertically integrated pulp and paper mill are presented. Results suggest that integrated forest products facilities drawing their raw material from sustainably managed forests can achieve a net positive carbon balance over the product cycle. The amount of net carbon sequestration attributable to such facilities depends upon a number of factors. The most critical of these include net forest growth, the method for allocating the growth in forest carbon among all of those harvesting from the drain area of a given facility, and the use and disposal patterns for the paper or wood products manufactured.     

Sensitivity of the terrestrial biosphere to climatic changes: impact on the carbon cycle

The biosphere is a major pool in the global carbon cycle; its response to climatic change is therefore of great importance. We developed a 5 degrees x 5 degrees longitude-latitude resolution model of the biosphere in which the global distributions of the major biospheric variables, i.e. the vegetation types and the main carbon pools and fluxes, are determined from climatic variables. We defined nine major broad vegetation types: perennial ice, desert and semi-desert, tundra, coniferous forest, temperate deciduous forest, grassland and shrubland, savannah, seasonal tropical forest and evergreen tropical forest. Their geographical repartition is parameterized using correlations between observed vegetation type, precipitation and biotemperature distributions. The model computes as a function of climate and vegetation type, the variables related to the continental biospheric carbon cycle, i.e. the carbon pools such as the phytomass, the litter and the soil organic carbon; and carbon fluxes such as net primary production, litter production and heterotrophic respiration. The modeled present-day biosphere is in good agreement with observation. The model is used to investigate the response of the terrestrial biosphere to climatic changes as predicted by different General Circulation Models (GCM). In particular, the impact on the biosphere of climatic conditions corresponding to the last glacial climate (LGM), 18 000 years ago, is investigated. Comparison with results from present-day climate simulations shows the high sensitivity of the geographical distribution of vegetation types and carbon content as well as biospheric trace gases emissions to climatic changes. The general trend for LGM compared to the present is an increase in low density vegetation types (tundra, desert, grassland) to the detriment of forested areas, in tropical as well as in other regions. Consequently, the biospheric activity (carbon fluxes and trace gases emissions) was reduced.     

Development of a protocol for monitoring status and trends in forest soil carbon at a national level

The national Forest Health Monitoring (FHM) program requires protocols for monitoring soil carbon contents. In a pilot study, 30 FHM plots loblolly shortleaf (Pinus taeda L./Pinus echinata Mill.) pine forests across Georgia were sampled by horizon and by depth increments. For total soil carbon, approximately 40% of the variance was between plots, 40% between subplots and 20% within subplots. Results by depth differed from those obtained by horizon primarily due to the rapid changes in carbon content from the top to the bottom of the A horizon. Published soil survey information overestimated bulk densities for these forest sites. The measurement of forest floor depths as a substitute to sampling did not provide reliable estimates of forest floor carbon. Precision of replicate samples was approximately 10-30% for field duplicates and 5-10% for laboratory duplicates. Based on national indicator evaluation criteria, sampling by depth using bulk density core samplers has been recommended for national implementation. Additional procedures are needed when sampling organic soils or soils with a high percentage of large rock fragments.     

Summary of efficacy evaluations using aerially applied Gypchek® against gypsy moth in the U.S.A.

Abstract

Gypchek^®, the gypsy moth (Lymantria dispar L.) nucleopolyhedrosis virus product, is manufactured by the United States Department of Agriculture (USDA) Forest Service, and Animal and Plant Health Inspection Service under controlled conditions in a laboratory strain of gypsy moth larvae. Gypchek was registered with the U.S. Environmental Protection Agency in 1978 as a general use pesticide to control gypsy moth. This product has been the subject of intense research and development targeted toward maximizing efficacy while minimizing the cost of production and application. The current Gypchek tank mix is applied at 1.25 × 10¹²occlusion bodies (OB's) per hectare for each of two applications (3‐days apart) at 18.7 litres/ha per application.     

Spatial variability of soil organic carbon in grasslands: implications for detecting change at different scales

Extensive data used to quantify broad soil C changes (without information about causation), coupled with intensive data used for attribution of changes to specific management practices, could form the basis of an efficient national grassland soil C monitoring network. Based on variability of extensive (USDA/NRCS pedon database) and intensive field-level soil C data, we evaluated the efficacy of future sample collection to detect changes in soil C in grasslands. Potential soil C changes at a range of spatial scales related to changes in grassland management can be verified (alpha=0.1) after 5 years with collection of 34, 224, 501 samples at the county, state, or national scales, respectively. Farm-level analysis indicates that equivalent numbers of cores and distinct groups of cores (microplots) results in lowest soil C coefficients of variation for a variety of ecosystems. Our results suggest that grassland soil C changes can be precisely quantified using current technology at scales ranging from farms to the entire nation.     

Soil carbon dynamics in cropland and rangeland

Most soils in the Midwestern USA have lost 30 to 50% of their original pool, or 25 to 40 Mg C/ha, upon conversion from natural to agricultural ecosystems. About 60 to 70% of the C thus depleted can be resequestered through adoption of recommended soil and crop management practices. These practices include conversion from plow till to no till, frequent use of winter cover crops in the rotation cycle, elimination of summer fallow, integrated nutrient management along with liberal use of biosolids and biological nitrogen fixation, precision farming to minimize losses and enhance fertilizer use efficiency, and use of improved varieties with ability to produce large root biomass with high content of lignin and suberin. The gross rate of soil organic carbon (SOC) sequestration ranges from 500 to 800 kg/ha/year in cold and humid regions and 100 to 300 kg/ha/year in dry and warm regions. The rate of SOC sequestration can be measured with procedures that are cost effective and credible at soil pedon level, landscape level, regional or national scale. In addition to SOC, there is also a large potential to sequester soil inorganic carbon (SIC) in arid and semi-arid regions. Soil C sequestration has numerous ancillary benefits. It is truly a win-win situation: extremely cost-effective, and a bridge to the future until alternative energy options take effect.     

Inventory of Emission for the City of Chicago

A review of the methodology used in the development and implementation of a comprehensive emission inventory survey with emphasis on manufacturing plants. This inventory, a major goal in the Dynamic Air Resource Management Program for the City of Chicago, is a principal objective of a Five-Year Technical Assistance Agreement with the United States Public Health Service, Division of Air Pollution and Chicago’s Department of Air Pollution Control. Information obtained through a 75 percent response from a questionnaire mailed to 7312 manufacturing firms provided the basis for this inventory. Data on the delinquent firms was obtained from on-site evaluation by department engineers. This report will also cover computer processing and tabulation of survey data.     

A greenhouse gas emissions inventory for Pennsylvania

The Pennsylvania greenhouse gas (GHG) emissions inventory presented in this paper provides detailed estimates of emissions and their sources for the six major categories of GHGs. The inventory was compiled using the current U.S. Environment Protection Agency methodology, which applies emissions factors to socioeconomic data, such as fossil energy use, vehicle miles traveled, and industrial production. The paper also contains an assessment of the methodology and suggestions for improving accounting with respect to process, sectoral, and geographic considerations. The study found that Pennsylvania emitted 77.4 million metric tons carbon equivalent of GHGs in 1990 and that this total increased by 3% to 79.8 million metric tons carbon equivalent by 1999. Despite this increase, however, the state's percentage contribution to the United States total declined during the decade. Pennsylvania's carbon dioxide (CO2) emissions from fossil fuels represented 92.4% of 1990 totals and declined to 90.5% in 1999. Electricity generation was the largest single source of CO2 emissions, being responsible for 38% of fossil fuel CO2 emissions in 1990 and 40% of the total in 1999. Transportation emissions accounted for the largest increases in emissions between 1990 and 1999, whereas industrial emissions accounted for the largest decrease. The overall trend indicates that Pennsylvania has been able to weaken the relationship between GHG emissions and economic growth.