Inventory methodologies

IPCC factor



The N2O emission factor for animal excreta deposited during grazing (EF3PRP)

The N2O emission factor for leached N (EF5)

Biological N fixation as a source of N2O (FBN)

The annual amount of N input from crop residue (FCR)

N sources from which the amount of leached N and subsequent indirect N2O emissions is estimated

Default value for sheep and 'other IPCC (2006); animals' reduced from 0.02 to 0.01 kg C. A. M. de Klein N2O-N (kg N)-1 Remains at 0.02kg (2004, unpublished) N2O-N (kg N)-1 for cattle (dairy, non-dairy and buffalo), poultry and pigs

Default value reduced from 0.025 to 0.0075kg N2O-N/per kg N lost in leachate or runoff water

Hiscock etal (2003); Dong et al (2004); Reay et al

(2004); Sawamoto et al

Removed from inventory as a direct Rochette and source of N2O. N2O emissions induced Janzen (2005) by the growth of legume crops/forages may be estimated solely as a function of the above- and below-ground N inputs from crop/forage residues

N mineralization associated with C loss due to land-use change and management practices (FSOM)

Amounts of applied mineral N fertilizers (FSN) and organic N fertilizers (FON)

Now includes the contribution of below-ground N to the total input of N from crop residues

Now also includes the N released following pasture renewal

Now included as a new source of N2O

The default methodology no longer adjusts Fsn and FON for the amounts of NH3 and NOx volatilized after application to soil before applying the N2O EF

Now includes N inputs from crop residue (FCR) and from N mineralization associated with C loss due to land-use change (FSOM)

IPCC (2006)

van der Weerden et al (1999); Davies et al (2001)

Smith and Conen (2004)

IPCC (2006)

IPCC (2006)

1996 IPCC guidelines. Recently, Rochette et al (2008) used the same regression approach as Bouwman (1996) and Bouwman et al (2002) to estimate the N2O emission factor for N fertilizer from Canadian field studies. Their results suggested average background emissions of 0.1, 0.6 and 0.8kg N2O-N ha-1 yr-1 for Canadian agricultural soils in three different regions.

Some countries have adapted the IPCC default methodology to include 'background' emissions. For example, in Sweden a 'background' EF of 0.5kg N2O-N/ha/year for mineral soils (Kasimir-Klemedtsson, 2001) is used. This value was determined on the basis of the country-specific EF for inorganic N fertilizer inputs of 0.8 per cent, which is lower than the IPCC default value of 1.25 per cent. The reason for including these 'background' emissions in the Swedish inventory is to account for the effect of N mineralization following decomposition of soil organic matter after cultivation events (Kasimir-Klemedtsson, 2001). This thus recognizes N fertilizer inputs and soil cultivation as two separate sources of N2O. As mentioned above, this was also recognized in the 2006 revision of the IPCC guidelines and cultivation effects are now included as a separate source in the 2006 IPCC guidelines. However, in addition to the cultivation effect, 'background' emissions from uncultivated improved land are higher than those from natural land, and these anthropogenic emissions are currently not accounted for in the default IPCC methodology. As a result, the New Zealand and Australian inventories do not include a correction for background N2O emissions. The Australian National Inventory Report further argues that, in contrast to European and North American agriculture, there has been little accumulation of soil nitrogen from previous cropping that might otherwise predispose these soils to substantial background N2O emission rates (DCC, 2008).

Farm-scale accounting

The IPCC methodology has been incorporated into on-farm inventory models to estimate N2O emissions at a farm scale. Examples of such models are the nutrient budgeting model OVERSEERĀ® (Wheeler et al, 2003, 2008); the Australian Farm Greenhouse Gas Accounting Tools (Eckard, 2008) and various whole-farm models such as DairyWise or Farmgreenhouse gas (Olesen et al, 2006; Schils et al, 2007; van Groenigen et al, 2008). These models generally operate at an annual time step.

The nutrient budgeting model OVERSEERĀ® (Wheeler et al, 2003) is extensively used throughout New Zealand as a tool for optimizing on-farm nutrient management as well as implementing resource requirements from local authorities to minimize N and P losses to waterways. The model includes a greenhouse gas module for estimating on-farm emissions of methane, nitrous oxide and carbon dioxide (Wheeler et al, 2008), and is based on the New Zealand IPCC (NZ-IPCC) inventory methodology (Ministry for the Environment, 2008). It thus estimates N2O emissions from the size of N inputs in the system multiplied by an EF for each input. The model can operate in two modes. One mode uses the NZ-IPCC values for the EFs and fractions for N leaching and ammonia volatilization. The other mode is more 'site-specific' and uses disaggregated EFs for animal excreta based on animal type and soil drainage class. This mode also uses the nutrient budgeting model to estimate N leaching and volatilization losses, instead of using the NZ-IPCC fractions. Both modes use farm-specific input variables on animal production, fertilizer use and pasture quality, and can account for N mitigation strategies such as optimized N fertilizer form and timing, nitrification inhibitor use, low N feed, and destocking or housing options (Wheeler et al, 2008).

Schils et al (2007) reviewed a range of whole-farm models for estimating greenhouse gas emissions from dairy systems, including DairyWise, Farm-greenhouse gas, SIMSdairy and FarmSim. These models often use IPCC inventory methodologies for estimating CH4 and N2O emissions (and to some extent CO2), but the authors concluded that they should not be seen as a replacement for the IPCC methodology, which is clearly aimed at providing transparent and consistent reports on national greenhouse gas emissions. The whole-farm models provide powerful tools for evaluating the impact of mitigation strategies at an individual farm level (Schils et al, 2007). In addition, these models are useful tools for examining the changes in IPCC EFs and fractions that would result from adoption of mitigation strategies, and can thus help to adjust the IPCC inventory methodology to account for these mitigations.

Biophysical models

Modelling provides a valuable complement to measurement, extending limited temporal and spatial measurements to other climatic and edaphic conditions, regions and scales. In recent years, the development of biophysical N2O emission models has received much attention and a large number of models have been developed to simulate N2O fluxes from natural and managed ecosystems (Dalal et al, 2003; Chen et al, 2008).

In contrast to the empirical inventory models, detailed biophysical, process-based, daily time-step models tend to simulate changes in environmental factors affecting the emissions (for example N transformations, moisture, temperature and available carbon) from key input variables (for example climate, soil type, system and management practices), thereby predicting the soil N pools available for nitrification and denitrification to N2O. These models apply largely mechanistic algorithms to estimate the effect of the proximal factors on N2O emissions and reflect our understanding of drivers of N2O emissions at a process level. Whole-farm systems models tend to use a combination of accounting, empirical and mechanistic modelling approaches based on daily, monthly or annual time steps.

Some highly mechanistic models, such as DNDC (Li et al, 1992a) and several others (Smith, 1980; Grant and Pattey, 1999; Riley and Matson, 2000) simulate microbial growth rates and solute and gas transport through the soil profile and aggregates. Other models simulate nitrification and denitrification as a function of frequently measured and modelled variables such as soil water, temperature, NO3~, NH4+ and soil respiration, for example DAYCENT (Del Grosso et al, 2006), EcoMod and DairyMod (Johnson et al, 2008) and WNMM (Li et al, 2007). More generalized models have been developed to simulate N2O fluxes at regional and global scales (Parton et al, 1996; Potter et al, 1996). The annual time-step models are more generally applicable but are surrounded by great uncertainties, while the more complex daily time-step models are highly site-, climate- and system-specific, often requiring significant parameterization.

The different N2O models described above have different purposes, from annual accounting to evaluating potential mitigation strategies and/or improving understanding of the drivers of N2O emissions at a process level. The level of complexity and the detail of the required input data generally increase with the extent to which a model accounts for soil, climate and management variability. Depending on the purpose of the model, a balance is often required between available input data and the reliability of the estimates. Although N2O emission models have been further developed and improved in recent years, the spatial and temporal variability in emissions and complex interactions between the driving variables challenge our ability to predict emissions (Calanca et al, 2007). For example, in a recent comparison between the process-based models DAYCENT and DNDC for estimating N2O emissions from cropping systems, Smith W. N. et al (2008) concluded that the DNDC model accurately predicted average seasonal N2O emissions, whereas DAYCENT underestimated them by up to 58 per cent. However, as neither model accurately simulated the timing of individual emission events, the authors concluded that improvements to the soil moisture and nitrogen transformation modules of the models are required before enhancements are made to the N2O emission routines. Another contributing factor affecting the simulation of the timing of N2O emissions is that most mechanistic models include algorithms describing N2O production in the soil, but do not include processes and time steps for the actual N2O emission from the soil surface.

0 0

Post a comment