Uncertainties in soil C inventories are related at Tiers 1 and 2 to representation of 1) land-use and management activities; 2) mineral soil reference C stocks; and 3) stock change and emission factors. Tier 3 uncertainties depend on model structure and parameters, or measurement error/sampling strategy. Uncertainty is generally reduced by more sampling and use of a higher Tier estimates incorporating country-specific information.
Uncertainties in reference C stocks and emission factors are indicated in Table 2.3 in Chapter 2; Tables 5.5 and 5.6 in Chapter 5; and Tables 6.2 and 6.3 in Chapter 6. Uncertainties in land-use and management data will need to be assessed by the inventory compiler, and combined with uncertainties for the default factors and reference C stocks using an appropriate method, such as simple error propagation equations. If using aggregate land-use area statistics for activity data (e.g., FAO data), the inventory compiler may have to apply a default level of uncertainty for the land area estimates (+50%). However, it is good practice for the inventory compiler to derive uncertainties from country-specific activity data instead of using a default level.
Default reference C stocks for mineral soils and emission factors for organic soils can have high uncertainties, when applied to specific countries. Defaults represent globally averaged values of land-use and management impacts or reference C stocks that may vary from region-specific values (Powers et al, 2004; Ogle et al, 2006). Bias can be reduced by deriving country-specific factors using Tier 2 method or by developing a Tier 3 country-specific estimation system. The underlying basis for higher Tier approaches will be research in the country or neighbouring regions that address the effect of land use and management on soil C. It is good practice to minimize bias by accounting for significant within-country differences in land-use and management impacts, such as variation among climate regions and/or soil types, even at the expense of reduced precision in the factor estimates (Ogle et al, 2006). Bias is more problematic for reporting stock changes because it is not necessarily captured in the uncertainty range (i.e., the true stock change may be outside of the reported uncertainty range if there is significant bias in the factors).
Precision in land-use activity statistics may be improved through a better national system, such as developing or extending a ground-based survey with additional sample locations and/or incorporating remote sensing to provide additional coverage. It is good practice to design a classification that captures the majority of land-use and management activity with a sufficient sample size to minimize uncertainty at the national scale.
For Tier 2 methods, country-specific information is incorporated into the inventory analysis for purposes of reducing bias. For example, Ogle et al. (2003) utilized country-specific data to construct probability density functions for US specific factors, activity data and reference C stocks for agricultural soils. It is good practice to evaluate dependencies among the factors, reference C stocks or land-use and management activity data. In particular, strong dependencies are common in land-use and management activity data because management practices tend to be correlated in time and space.
Tier 3 models are more complex and simple error propagation equations may not be effective at quantifying the associated uncertainty in resulting estimates. Monte Carlo analyses are possible (Smith and Heath, 2001), but can be difficult to implement if the model has many parameters (some models can have several hundred parameters) because joint probability density functions must be constructed quantifying the variance as well as covariance among the parameters. Other methods are also available such as empirically-based approaches (Monte et al., 1996), which use measurements from a monitoring network to statistically evaluate the relationship between measured and modelled results (Falloon and Smith, 2003). In contrast to modelling, uncertainties in measurement-based Tier 3 inventories can be determined directly from the sample variance, measurement error and other relevant sources of uncertainty.
Was this article helpful?
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.