Measurementbased Tier 3 inventories

Inventories can be based on direct measurements of C stock changes from which emissions and removals of carbon are estimated. Measurement of some non-CO2 greenhouse gas emissions is possible, but because of the high spatial and temporal variability of non-CO2 emissions, Tier 3 methods will likely combine process models with measurements to estimate non-CO2 emissions. Purely measurement-based inventories, e.g., based on repeated measurements using a national forest inventory can derive carbon stock change estimates without relying on process models, but they do require appropriate statistical models for the spatial and temporal scaling of plot measurements to a national inventory. Approaches based on dynamic models (e.g., process-based models) to estimate national emissions will be discussed in Section 2.5.2. In general, six steps are involved with implementation of a Tier 3 measurement-based inventory.

Step 1. Develop sampling scheme. Sampling schemes can be developed using a variety of approaches, but typically involve some level of randomization of sampling sites within strata. (Even inventories based on a regular grid typically select the starting point of the grid at random). Inventory compilers will determine an appropriate approach given the size of their country, key environmental variables (e.g., climate) and management systems in their region. The latter two may serve as stratification variables, assuming the sampling scheme is not completely random. In addition, it is good practice for sampling to provide wide spatial coverage of emissions and/or removals for a particular key source category.

The inventory compiler should establish an appropriate time period over which sites will be re-sampled if using a repeated measures design. The timing of re-measurement will depend on the rate of stock changes or non-CO2 greenhouse gas emissions. For example, re-measurement periods in boreal and some temperate regions, where trees grow slowly and DOM pools change little in single years, can be longer than in environments where carbon dynamics are more rapid. Where fluxes are measured directly, greater temporal and spatial variability will require more frequent or more intensive sampling to capture fluxes which might otherwise be missing from the measurement record.

Some approaches do not include re-sampling of the same sites. Such designs are acceptable, but may limit the statistical power of the analysis, and therefore lead to greater uncertainty. It is likely that a repeated measures design will provide a better basis for estimating carbon stock changes or emissions in most countries.

It is good practice to develop a methodology handbook explaining the sampling scheme as part of Step 1. This handbook can be useful for those involved with the measurements, laboratory analyses and other aspects of the process, as well as possibly providing supporting material for documentation purposes.

Step 2. Select sampling sites. Specific sampling sites will be located based on sampling design. It is good practice to have alternative sites for sampling in case it is not possible to sample some original locations. In a repeated measures design, the sites will become a monitoring network that is periodically re-sampled.

Determining sampling locations will likely involve the use of a geographic information system. A geographic database may include a variety of environmental and management data, such as climate, soils, land use, and livestock operations, depending on the source category and stratification. If key data are not available at the national scale, the inventory developer should re-evaluate the design and stratification (if used) in Step 1 and possibly modify the sampling design.

Sampling may require coordination among different national ministries, provincial or state governments, corporate and private land owners. Establishing relationships among these stakeholders can be undertaken before collecting initial samples. Informing stakeholders about ongoing monitoring may also be helpful and lead to greater success in implementing monitoring programs.

Step 3. Collect initial samples. Once the final set of sites are determined, a sampling team can visit those locations, establish plots and collect initial samples. The initial samples will provide initial carbon stocks, or serve as the first measure of emissions. It is good practice to establish field measurement and laboratory protocols before the samples are collected. In addition, it may be helpful to take geographic coordinates of plot locations or sample points with a global positioning system, and, if repeated measures are planned, to permanently mark the location for ease of finding and re-sampling the site in the future.

It is good practice to take relevant measurements and notes of the environmental conditions and management at the site. This will confirm that the conditions were consistent with the design of the sampling scheme, and also may be used in data analysis (Step 5). If a stratified sampling approach is used, and it becomes apparent that many or most sites are not consistent with the expected environmental conditions and management systems, it is good practice to repeat Step 1, re-evaluating and possibly modifying the sampling scheme based on the new information.

Step 4. Re-sample the monitoring network on a periodic basis. For repeated measures designs, sampling sites will be periodically re-sampled in order to evaluate trends in carbon stocks or non-CO2 emissions over an inventory time period. The time between re-measurement will depend on the rate of stock changes or the variability in emissions, the resources available for the monitoring program, and the design of the sampling scheme.

If destructive sampling is involved, such as removing a soil core or biomass sample, it is good practice to resample at the same site but not at the exact location in which the sample was removed during the past. Destructive sampling the exact location is likely to create bias in the measurements. Such biases would compromise the monitoring and produce results that are not representative of national trends.

Step 5. Analyze data and determine carbon stock changes/non-CO2 emissions, and infer national emissions and removal estimates and measures of uncertainty. It is good practice to select an appropriate statistical method for data analysis based on the sampling design. The overall result of the statistical analysis will be estimates of carbon stock changes or measurements of emissions from which the national emission and removal estimates can be derived. It is good practice to also include estimates of uncertainty, which will include measurement errors in the sample collection and laboratory processing (i.e., the latter may be addressed using standards and through cross-checking results with independent labs), sampling variance associated with monitoring design and other relevant sources of uncertainty (see discussion for each source category later in this volume in addition to the uncertainty chapter in Volume 1). The analysis may include scaling of measurements to a larger spatial or temporal domain, which again will depend on the design of the sampling scheme. Scaling may range from simple averaging or weighted averaging to more detailed interpolation/extrapolation techniques.

To obtain national estimates of stock changes or emission of non-CO2 greenhouse gases, it is often necessary to extrapolate measurements using models that take into consideration environmental conditions, management and other activity data. While the net changes of carbon-based greenhouse gasses can (at least in theory) be estimated purely by repeated measurements of carbon stocks, statistical and other models are often employed to assist in the scaling of plot measures to national estimates. National emission estimates of non-CO2 greenhouse gases are unlikely to be derived from measurements alone because of the expense and difficulty in obtaining the measurement. For example, N2O emissions from forest fires cannot be measured empirically but are typically inferred from samples, activity data on the area burnt, and fuel consumption estimates. In contrast, soil N2O emissions can be readily estimated using chambers, but it would be very expensive to establish a network with the sampling intensity needed to provide national emission estimates based solely on measurements without use of models for extrapolation.

It is good practice to analyze emissions relative to environmental conditions in addition to the contribution of various management practices to those trends. Interpretation of the patterns will be useful in evaluating possibilities for future mitigation.

Step 6. Reporting and Documentation. It is good practice to assemble inventory results in a systematic and transparent manner for reporting purposes. Documentation may include a description of the sampling scheme and statistical methods, sampling schedule (including re-sampling), stock change and emissions estimates and the interpretation of emission trends (e.g., contributions of management activities). In addition, QA/QC should be completed and documented in the report, including quality assurance procedures in which peer-reviewers not involved with the analysis evaluate the methodology. For details on QA/QC, reporting and documentation, see the section dealing with the specific source category later in this volume, as well as information provided in Volume 1, Chapter 6.

Was this article helpful?

0 0
Guide to Alternative Fuels

Guide to Alternative Fuels

Your Alternative Fuel Solution for Saving Money, Reducing Oil Dependency, and Helping the Planet. Ethanol is an alternative to gasoline. The use of ethanol has been demonstrated to reduce greenhouse emissions slightly as compared to gasoline. Through this ebook, you are going to learn what you will need to know why choosing an alternative fuel may benefit you and your future.

Get My Free Ebook

Post a comment