Voluntary Greenhouse Gas Reporting Feasibility Study
3 Estimating GHG emissions in NZ
3.1 Introduction
This section contains an assessment of methods that could be used to estimate GHG emissions at a farm scale in New Zealand.
There is no existing tool or system available in New Zealand to specifically estimate GHG emissions at the farm scale. New Zealand follows a methodology to estimate GHG emissions at a national level as part of its commitment to the United Nations Framework Convention on Climate Change. This methodology could be adapted to use at the farm scale to provide the methodology behind the VGGR system.
Farm scale models exist for other on farm purposes (e.g. the Overseer® model used for nutrient budgeting) that could be used to predict GHG emissions. Some of these models adopt the national inventory methodology as a basis for estimating emissions, while others contain different, process based methodologies. The government could choose to adapt one of these farm scale models to provide both the methodology and the interface of a VGGR, or could choose to create a new model and interface based on the national inventory method.
This section contains an overview of the national inventory methodology for both CH4 and N2O and contains an assessment of the value of adopting this methodology at a farm scale for the VGGR.
This section then provides an overview of the existing farm scale tools available and assesses, using a number of predetermined criteria, the advantages and disadvantages of adapting an existing model, or creating a new model to provide the basis of the VGGR. Recommendations are then provided in relation to development of a VGGR to estimate GHG emissions at the farm scale in New Zealand.
3.2 National inventory method
As a signatory to the United Nations Framework Convention on Climate Change (UNFCCC) and the Kyoto Protocol, New Zealand is obliged to compile annual inventories of its GHG emissions. These inventories are externally reviewed to assess whether they comply with international best practice. To help countries develop robust GHG inventories the Intergovernmental Panel on Climate Change (IPCC) publishes guidelines and methodologies that can be followed or adapted to suit national circumstances (IPCC 2000, IPCC 1996). IPCC guidelines are separated into two broad categories; Tier 1 inventories and Tier 2 inventories. Tier 1 inventories are relatively simple inventories that require less data and rely heavily on the use of published default values. Tier 2 inventories have greater data requirements, and use more complex methodologies to estimate emissions for individual circumstances rather than relying on default values. Developed countries such as New Zealand are expected to use Tier 2 inventories wherever possible, especially for sectors where GHG emissions make up a major component of the national emissions. In New Zealand, because of the importance of agricultural emissions to the national emissions profile Tier 2 methods have been developed for both agricultural CH4 and N2O inventories.
National emissions estimated using IPCC methodologies are top down estimates that use data aggregated at the national scale. They therefore provide no information at the farm scale; although using default emission factors (e.g. emissions per head of stock) it is possible to estimate individual farm emissions. These estimates however will be differentiated at the farm level by animal population only rather than the broad range of individual circumstances that will influence emissions in practice. To date in New Zealand there has been no concerted effort to estimate national emissions from the farm scale up although some models can be used to estimate emissions at the farm scale.
In any discussion on the methods available for estimating agricultural GHG emissions at any scale the following issues are relevant:
• CH4 & N2O emissions from agriculture can only be measured experimentally at the individual animal, sub herd or paddock scale – emissions at farm, regional or national scale are estimated
• emissions are not constant over time – CH4 & N2O emissions vary hourly, daily, weekly, monthly & annually
• emissions vary in space – patch, paddock, farm & region
• there are multiple influences on emissions – environmental, physical and biological
• the processes influencing emissions are not fully understood.
The above points mean that when deciding on an appropriate method for on-farm estimates the emissions model complexity will have to be balanced with the data required to drive the model. In addition, uncertainties surrounding any estimates are likely to be large.
3.2.1 CH4 national inventory method
Details of the methods adopted in New Zealand for estimating CH4 emissions are outlined briefly in the annual national inventory report (MfE 2006); full details for enteric CH4 can be found in Clark et al (2002) and for emissions from manure management in Saggar et al (2004). Only brief details will be given here. A schematic of the current CH4 inventory method is presented in Figure 6-2.
The methodology adopted involves three basic steps
1) Detailed categorisation of animal populations using data from the Ministry of Agriculture and Forestry (MAF) annual census/survey. These population models break down the population into species (e.g. dairy cattle, non-dairy cattle, sheep etc) and sub-categories within each species (e.g. breeding ewes, breeding rams, lambs etc) on a monthly basis.
2) Estimation of the quantity of feed eaten by the ‘average’ animal in each species and sub-category on a monthly basis. This is done by estimating the energy requirements of the average animal using algorithms developed by CSIRO in Australia (SRC 1990). The data needed for these estimations includes animal size, animal age, animal productivity (milk yield, growth rate), and diet quality. Energy requirements can be converted into the quantity of feed eaten from a knowledge of the energy concentration of the ingested diet.
3) Estimation of the quantity of CH4 emitted per unit of feed ingested. These values are obtained from measurements made on representative groups of animals in New Zealand.
When generating values for the CH4 national inventory methodology the following approach is adopted:
• emissions from each sub-category each month are estimated from consumed x CH4 emitted /unit feed x population number.
• total emissions for each sub-category are the sum of monthly emissions
• emissions for each species are the summation of annual emissions from each sub-category; and
• national emissions are the summation of annual emissions from each species.
The only exceptions to the above methodology are emissions from minor species (goats, horses, pigs and poultry) where IPCC or New Zealand derived default values are used.
Emissions from manure management involve steps 1 and 2 above. After estimating the quantity of feed eaten by each sub-category of animal, the quantity of faecal material produced is estimated from data on diet digestibility. Emissions from faecal material are then determined using a mixture of New Zealand and internationally published data. As with enteric CH4 emissions IPCC default values are used for non-ruminant animals.
The data requirements of the national inventory method are modest although even then data are not readily available in all circumstances and some estimation is required. The data requirements and sources are summarised in Table 3-1. All data are needed at the national level only.
Table 3-1 Data needs and sources for national CH4 inventory
| Data needed | Dairy cattle | Beef cattle | Sheep | Deer |
|---|---|---|---|---|
| Annual population by livestock sub-category | MAF livestock statistics | MAF livestock statistics | MAF livestock statistics | MAF livestock statistics |
| Live weight | Livestock Improvement Corporation (LIC) survey | Estimated from MAF slaughter statistics | Estimated from MAF slaughter statistics | Estimated from MAF slaughter statistics |
| Animal Productivity | LIC Dairy statistics | Estimated from MAF slaughter statistics | Estimated from MAF slaughter statistics Wool yields from Meat and Wool New Zealand (MWNZ) | Estimated from MAF slaughter statistics |
| Monthly diet quality (digestibility, metabolisable energy), N %) | Farm survey, assumed not to vary with time | Farm survey, assumed not to vary with time | Farm survey, assumed not to vary with time | Farm survey, assumed not to vary with time |
3.2.2 N2O national inventory method
The two main pathways for N2O emissions from New Zealand agriculture are those arising from synthetic N applications and those arising from the direct deposition of animal wastes onto pastures. The methods used to estimate emissions arising from these pathways are briefly described in the annual national inventory report (MfE 2006) and full details can be found in the IPCC’s report Good Practice Guidance and Uncertainty Management in National Greenhouse Gas Inventories (IPCC 2000). Only a brief description will be given here. A schematic of the current N2O inventory method is presented in
Figure 6-3.
The methodology for N2O estimates from synthetic N applications and the deposition of animal wastes has two basic steps.
1) An estimation of the quantity of N deposited onto pastures. For synthetic N fertilisers, data on the total quantity of fertiliser sold in New Zealand is obtained from the Fertiliser Manufacturers Research Association (FMRA). There is no attempt to break down fertiliser usage on a sectoral basis. To estimate animal deposition of N, data on the quantity of feed eaten are combined with data on the N content of feed and an estimate of the amount of N retained in animal products (meat and milk); N consumed – N retained = N deposited onto pastures. The estimation of feed intake follows exactly the same methodology as described for CH4 above. N deposition by pigs and poultry is estimated using IPCC default values.
The estimation of direct and indirect emissions arising from the deposited N. Irrespective of the source of N deposited on pastures a similar methodology, albeit with slightly different parameter values, is used to estimate N2O emissions. Direct emissions are simply a fixed proportion of the N applied/deposited. Indirect emissions of N2O arise from two sources, volatilised ammonia (NH3) and nitrogen oxides (NOx) which are returned to soils during rainfall and then re-emitted as N2O; and leaching/run off of N which enters water systems and emits N2O on its movement to the sea. Fixed proportions of the deposited/applied N are assumed to be lost to volatilisation and leaching, and fixed quantities of N2O arise from leached and volatilised N. The parameter values used to estimate emissions are a mixture of IPCC defaults and New Zealand specific values.
Since the estimation of N2O emissions requires an estimate of herbage intake as the first step in calculating N intake and retention, the data needs are exactly the same as the CH4 inventory up until the point that herbage intake is calculated. The only additional data needed to complete the calculations are the N% of the ingested diet (Table 3-1).
3.3 Value of adopting the national inventory for VGGR
The advantages and disadvantages of using the national methodology at the farm scale are presented in Table 3-2.
Table 3-2 Advantages/disadvantages of using the national inventory methodologies at a local scale
| Advantages | Disadvantages |
| Methodology well developed & relatively simple | A national ‘average’ non-process based method that doesn’t capture individual circumstances well, especially for N2O |
| Uses existing data and/or data available to individual farmers | Some mitigation options difficult to fit into methodology |
| Some ability to incorporate mitigation technologies | Not as accurate as process based approaches because some important driving variables (e.g. climate and soil type) are not included and doesn’t capture range of managements practised on farms |
| Individual farm data can be used to improve national methodology/emission estimates | |
| Methodology accepted by international monitoring agencies (UNFCCC) & is fixed for first commitment period of Kyoto Protocol |
From a purely technical perspective the current national CH4 emission methodology does have the ability to incorporate the principal drivers of CH4 emissions if it is used at a farm scale. This is basically because animal numbers/type and feed intake are the biggest influences on CH4 production. Diet quality influences are partially taken into account during the estimation of intake, although the direct effect of changes in diet on CH4 emissions per unit of intake are ignored if constant values are used for CH4/kg DMI. Any attempts to develop a more process based approach to predicting CH4 emissions from a detailed characterisation of feeds are hindered by the complexity of the data needed to run process based models and the lack of direct cause/effect understanding between feedtype and CH4 emissions. . A lack of data from animals grazing fresh forage diets rules out at this stage the use of less empirical approaches to predicting CH4 emissions from the wide variety of feeds and feeding conditions encountered in practice. Taking into account the status of technical knowledge the current national inventory approach to estimating CH4 emissions would appear to be suitable for use in an on-farm recording system.
Predicting N2O emissions is complex because of the large number of variables influencing the processes of N2O formation. The present national methodology does not attempt to capture many of these complexities (e.g. soil type and climate) and hence some of the major influences on emissions are not taken into account. When estimating emissions at the farm scale it would be preferable if these local influences could be captured in the methodology, although a more complex approach has greater data needs and these data may not be always readily available. Choosing a more appropriate methodology for farm scale use will therefore have to balance simplicity with accuracy of prediction. Developing a more appropriate farm based method of recording will also provide excellent data at the national level for testing the appropriateness of the national estimation method.
3.4 Existing farm scale models that could be used to estimate GHG emissions
Although there have been no systematic attempts to calculate emissions at the individual farm scale in
New Zealand there are several tools/models available that could possibly be used in a Voluntary Greenhouse Gas Reporting (VGGR) system. These range from simple approaches using derived emission factors that need minimum on-farm data, to complex process based models that are, judged from a farm perspective, data hungry. While only a few models have been used to model agricultural GHG emissions there are many more models that could estimate emissions with relatively small modifications, especially if IPCC methodologies are adopted for farm scale emission estimates. For example, any model that can estimate intake from animal performance and feed quality can predict CH4 emissions per animal if a fixed factor is used for the quantity of CH4 emitted per unit of intake. The addition of data on the N% in the feed will allow for the estimation of N2O emissions. The only methods considered in this report are New Zealand methods. Scientists throughout the world have been modelling agricultural GHG emissions at a range of scales but, for these to be useful in New Zealand, they have to capture the specific New Zealand farm circumstances and this is almost certain to mean extensive and time consuming modification. For example DNDC, a model used by Landcare scientists in New Zealand is used in both North America and the UK to predict N2O emissions but has had to be substantially modified over a number of years for it to be suitable for use here.
3.4.1 Derived emission factors for CH4 & N2O
New Zealand uses country specific Tier 2 methods for estimating CH4 and N2O emissions and does not rely on published emission factors. However, CH4 emission factors can be derived for New Zealand by dividing national estimates of total emissions for each animal species and species sub-category by annual population figures. Similar data can be derived for N2O emissions. Derived emission factors are available for sub-categories of the principle rumen species by age and gender and these could be used to estimate annual emissions at the farm scale very simply. The only farm specific data needed by a farmer would be annual population data broken down by species and sub-category.
The national inventory methodology attempts to model the ‘average’ situation and does not refer to any specific situation or farming system and so nationally derived emission factors have limited applicability in any given situation and will result in large errors in estimated emissions at the farm scale. The only influence a particular farm business has on emissions calculated using derived emissions factors is the species population/type. In addition, using derived emission factors makes it very difficult to incorporate mitigation technologies in an on-farm recording scheme. From a farmer perspective the use of derived emission factors does not provide data that is helpful in estimating emissions from a particular management regime or in managing farm GHG emissions. From a government perspective it does not provide any new information on national emissions. As a general principle it would seem preferable for emissions to be estimated independently at the farm scale and built up to provide a national picture, rather than national estimates determining individual farm emissions.
Methods being used to predict CH4
DNDC4 is a soil process model used by Landcare that has a grass growth component from which herbage intake and hence CH4 emissions can be estimated. Environmental variables such as rainfall, temperature and solar radiation are used to predict grass growth and a fixed proportion of the herbage grown (70%) is assumed to be eaten; CH4 emission is a fixed proportion of the quantity of grass consumed (CH4/kg DMI taken from the national inventory). This approach is very simple and requires little farm scale data since climate data at local scales are available from published sources. This method can provide an alternative method at a national scale and give regional estimates of emissions. However, it is unsuitable for a farm scale emission estimation method since it does not take any account of individual farm circumstances (e.g. mix of stock, animal performance, quality of herbage, management expertise) and does not allow the incorporation of mitigation technologies. It does not predict CH4 emissions from faecal material deposited onto pastures.
Overseer®5 is a soil nutrient balance model with an animal component that is extensively used as a nutrient budgeting model. Individual animal herbage intakes (cattle & sheep) are estimated using performance data and energetic algorithms in a manner analogous to that used in the national inventory method. These intake data are then combined with the CH4/kg DMI figures used in the national inventory to predict CH4 emissions. CH4 emissions from pasture deposited faecal material are estimated as per the national inventory model. Estimated in this manner the data requirements at the farm level mirror the data required when estimating the national CH4 inventory; animal populations (monthly), live weight gain/milk yield, live weight and herbage quality (monthly). Of these, herbage quality may be the most difficult to obtain in many farm situations. Estimating emissions using the Overseer approach has the advantage that there is considerable scope for individual farm actions to be reflected in the farm emission inventory. The summing of individual farm emissions can be used to validate not only national emission estimates but the data used in the national emission estimates.
The Dexcel Whole Farm Model6 is a comprehensive farm simulation model for use in the dairy industry. It has the flexibility to operate at a relatively simple or complex level (e.g. it has several ways of estimating animal intake) and can predict intake, and hence CH4 emissions using CH4/kg DMI constants. It can also predict enteric CH4 emissions from first principles using detailed information on the intake of different dietary components. It also has all the information needed to predict CH4 emissions from faecal material using the methods employed in the current New Zealand national inventory. The principal drawback of the model is that it works only for dairy cattle. In addition, because of the complexity of some of the processes, it requires data that may not be readily available on farm (e.g. detailed chemical characterisation of feed). This is especially true if trying to predict CH4 emissions from first principles. The model in its current form is perhaps better characterised as a research model rather than a model that can be used on-farm.
EcoMod7 is a linked soil/animal/plant model that can predict CH4 emissions at the farm scale with data requirements similar to those used in the national inventory. The model estimates intake from animal performance and diet quality in a similar manner to the national inventory and uses CH4/kg DMI taken from the national inventory method to predict CH4 emissions. It also has all the information needed to predict CH4 emissions from faecal material using the methods employed in the current New Zealand national inventory. The model works for sheep and cattle but its use in a farm scale emission exercise is limited by its ability to handle a single class of animal only (e.g. milking dairy cows) rather than the mix of classes (milking, dry, mixed age) that are found on farms. The model in its current form is perhaps better characterised as a research model rather than a model that can be used on-farm.
3.4.2 Methods being used to predict N2O
DNDC, a process based model for estimating N2O emissions, has been used by Landcare scientists to estimate N2O emissions at a farm and regional scale (Giltrap et al 2006). The model uses national inventory data to predict N input to soils or assumes that a constant proportion of N intake (estimated from grass growth and utilisation) is retained by animals. The model computes on a daily time step using daily temperature and rainfall data but uses average daily N input values (annual total/365). The process based nature of the model makes it attractive for use at a farm scale because it can capture the individual circumstances that influence N2O emissions. The model does however require data that may not be available at the farm level scale (e.g. clay content of soil, organic carbon content, daily temperature, rainfall and solar radiation) although national databases may be able to supply some of these data at a sufficient level of detail. The single biggest drawback of the model is that it has no animal component and hence has limited ability to predict at the individual farm level the quantity and timing of N deposited onto pastures. The quantity and timing of N deposition are crucial to emission estimates derived from a process based model.
Overseer® predicts N2O emissions in two ways. Firstly, it can mimic the methodology used in the national inventory (in fact some constants used in the national inventory methodology are taken from Overseer®); animal intake is combined with dietary N% data to estimate soil N deposition and N2O is estimated using the algorithms presented in Figure 6-2. The model can also use a modified approach that utilises individual farm soil information to adapt emission estimates to local circumstances. The data requirements using the IPCC approach are the same as those for the national inventory and, as has already been discussed, should be available on many farms. The modified IPCC approach is more data hungry as it does require some individual farm soil and climatic information. Overseer® would appear to provide an option that is somewhere between the full process based approach of DNDC and the national inventory methodology. This option allows for individual farm circumstances to be better reflected in emission estimates without excessive additional data demands.
EcoMod also has a soil process module that can predict N2O emissions at a farm scale. N input to the soil, based on animal intake estimated from animal performance, is used to drive a detailed soil N model of which N2O is one of the outputs. At this stage, although the model can estimate N2O emissions, the accuracy of the predictions is still being tested. As with DNDC the data required to run the soil module make it difficult to envisage EcoMod being used as a farm prediction model.
Value of adapting an existing farm scale model or creating a new model
This section contains an assessment of the value of adapting one of the existing farm scale models (described in Section 3.4) to provide both the methodology and the interface of a VGGR, or, alternatively creating a new model based on the national inventory method. Section 3.5.1 sets out a set of criteria for assessment of each option. Section 3.5.2 contains a table providing a summary assessment for each option against the set criteria.
3.4.3 Criteria for assessment
For a VGGR system to have any chance of success it is likely to need to meet a minimum set of requirements.
From a government perspective it requires;
• an ability to predict CH4 and N2O with some accuracy;
• a limited number of modifications
• compatibility with existing methods to develop the National Inventory
• an ability to use data collected at the local scale to obtain national values
• legal robustness (if a VGGR is later used as part of a mandatory system involving rewards/penalties)
• international acceptance of the method, including compatibility with IPCC methods for the first commitment period of the Kyoto protocol.
From a farmer perspective it needs to;
• provide information on the effectiveness of GHG mitigation strategies – to allow the farmer to cost-effectively minimise GHG emissions
• be simple to access and user friendly
• utilise existing data wherever possible and not require the collection of additional data unless absolutely necessary
• link if possible to other farm-based calculations e.g. nutrient budgeting to provide a co-benefit
• allow individual actions (e.g. mitigation) to be reflected in estimated farm GHG emissions
• be a well tested methodology that will not be continually revised due to methodological improvements
Table 3-3 contains an assessment of the suitability of (1) adapting any one of the four existing models or (2) creating a new model based on the national inventory method. The assessment is completed against the criteria outlined in 3.5.1 above.
Table 3-3 Assessment of options to adapt model or create new model
| Criteria | Options | ||||
|---|---|---|---|---|---|
| Adapt an existing model | Develop new model (based on inventory method) | ||||
| DNDC | Overseer | Dexcel Whole Farm Model | EcoMod | ||
| General characteristics | Strong on soil processes but very weak on N inputs to the soil and has no animal component. | Widespread use as a nutrient budgeting model. Has both an animal and soil component. | Model ‘platform’ that can incorporate a variety of models. Very strong animal component (dairy only) but no soil component. |
Has a mechanistic soil model and an animal performance intake model. | N/A |
| Ability to predict N2O and CH4 | Predicts CH4, N2O predictions still being refined. | Predicts CH4 & N2O. | Can predict CH4 by more than one method. Doesn’t predict N2O. |
N2O predictions still being refined, CH4 predicted using NZ inventory method. | A computerised method already exists at a national scale for CH4 & N2O and can be easily modified to work at a farm scale. Limited ability for national N2O methodology, to incorporate determinants of emissions at farm scale. |
| Modifications required | Animal component that can estimate DMI and N inputs to the soil. Needs to be made more user friendly. |
Animal component algorithms need updating to fit in with current national inventory method. Needs to be made more user friendly. |
Beef & sheep capability + N2O capability. Needs to be made more user friendly. | Capability to have more than a single class of animal essential. Needs to be made more user friendly. |
N/A |
| Compatibility with national inventory methodology | Does not use national inventory methodology. | Predicts CH4 & N2O using national inventory methodology & has capacity to go beyond IPCC method for N2O. | Can use national inventory methodology for CH4, does not predict N2O. | CH4 predicted using NZ inventory method, N2O predictions still being refined. | Provides potential for validation of national inventory. |
| Ability to use local data | Not good for CH4, good for N2O. | Good for CH4, not so good for N2O. | Good for CH4. Does not estimate N2O. | Good for CH4, good for N2O. | Good for CH4, does not capture N2O drivers well. |
| Legal robustness | Complexity of inputs may remove legal robustness. | Yes as uses national inventory method. | Complexity of inputs may remove legal robustness. Doesn’t use inventory method. | Yes as uses national inventory method. | Yes as uses national inventory method. |
| Simple & user friendly | Research model that requires an expert user. | Model designed for use by researchers & farmers, but not user friendly. | A research model up to now but a decision support version planned. | Research model only that requires an expert user. | Simple dedicated model that can be tailored to needs of users. Default or individual values can be used. |
| Good use of existing data | High data requirement e.g. organic carbon, clay content, daily temperature, rainfall & solar radiation but these may be obtainable at a local scale from national databases. | Data requirements for GHG emissions modest. | Detailed chemical characterisation of feed required. | High data requirement as it runs on a paddock scale with a daily time step e.g. soil parameters, daily temperature, rainfall & solar radiation. | Most data needs expected to be available at the farm level. Can include default values if required. |
| Co-benefits | Co-benefits as it can predict N leaching, CO2 respiration & NH3 volatilisation. | Widespread use as a nutrient budgeting model. | No | Can predict N leaching, CO2 respiration & NH3 volatilisation. | Sole use, could be adapted to include nutrient budgeting. |
| Reflects individual actions | Not well for CH4 | Yes | Yes for CH4 | Yes for CH4 | Yes |
| Stable estimates | Good for CH4. Not good for N2O. |
Yes | Good for CH4. Not good for N2O. |
Good for CH4, N2O still being developed. |
Yes |
| Issues | Process based approach. Work needed to interface an animal model. Complex to operate. |
Requires simplification. |
Complex model limited to the dairy sector. Adding N2O capability simple if IPCC method used. Does not mimic the national inventory method.
|
Process based approach. Complex.
Does not use national inventory |
Requires creation of a new model. Not good on inclusion of farm scale N2O determinants. |
| Suitability for VGGR | Not suitable for VGGR system due to lack of an animal model and heavy data needs. | Yes, but would require modification if centralised and simplification. | Not suitable in its current form unless it is used in conjunction with a separate beef & sheep model. | Not suitable for VGGR system due to heavy data demands, lack of verification of N2O module and limitations of animal categorisation. | Appropriate if the scheme operates centrally for both data storage and GHG computation since a new model would be needed in this option. Appropriate if strong desire to use national inventory. The scheme could operate centrally at the processing level & the data storage level or operate centrally for data storage but locally for the calculations. |
3.5 Summary
Agricultural GHG emissions cannot be measured at the farm scale, they have to be estimated. The multiple influences on emissions and, in some areas, an incomplete understanding of the processes involved means that any emission estimates are subject to large uncertainties. Choice of method will have to balance complexity with data requirements.
Emission factors derived from national emission estimates can be used on-farm but they are unsatisfactory since they offer little helpful information either to the farmer or to government.
New Zealand has a well developed national emissions methodology that can, if desired, form the basis of a farm scale recording system. Adapting the current national inventory model for use at a farm scale is a good option for CH4 given that the current method has the ability to incorporate individual management actions. Adapting the methodology for N2O may be less suitable for because of the simplicity of the methodology approached compared to the complexity and multiple drivers of N2O emissions.
Overseer seems to be the most feasible existing model option for use at present since it has an animal component that is similar to the national inventory methodology and an N2O prediction routine that can go beyond the national inventory. It is also being used on farms now as a nutrient budgeting tool. DNDC is hampered by the lack of an animal component. The Dexcel Whole Farm Dairy Model has no soil routine and is restricted to dairy cattle only. EcoMod has both an animal and soil component but is restricted to single animal classes only and is more of a research model than an on-farm tool.
Any model that can predict feed intake from data on animal size and performance can be used in a VGGR if the national inventory approach is used. However, if the national inventory approach is to be used it would seem preferable to use the existing inventory methodology as the methods used have been subjected to international scrutiny. The scheme could operate centrally at the processing level & the data storage level or operate centrally for data storage but locally for the calculations.
To provide a farm scale GHG emissions estimation method for the VGGR, consideration should be given to development of a new model based on the national inventory methodology or adaptation of the Overseer model. Both options are based on the national inventory method, make good use of existing data and could provide a simple interface.
Overseer is already being used by farmers for nutrient budgeting therefore provides a co-benefit to farmers and would offer a familiar interface. Overseer provides good estimates for N2O but would require significant redevelopment to incorporate a central calculation model and database, and also updating to include the new animal routines used in the national inventory, and to simplify the interface.
A new model based on the national inventory method could provide the simplest option of enabling submission of data to a central calculation model and database, and could also be developed to enable calculation at the farm level. A new model would require inclusion of a more complex approach to estimate N2O emissions and incorporation of a range of currently available mitigation practices. This would require inclusion of data entry for local factors such as soil and climate at the farm level and use of nitrification inhibitors. As a new model based on the national inventory method is likely to be the simplest option of providing a VGGR system, it is also likely to provide the most economic option for government.
4 DeNitrification DeComposition (Li et al., 1992)
5 Overseer is a registered trademark of AgResearch. The software and its output are copyrighted to AgResearch Ltd 2005.
6 Dexcel’s Whole Farm Model (Beukes et al., 2004)
7 EcoMod, is funded by AgResearch, NZ
Contact for Enquiries
Sustainable Land Management and Climate Change
MAF
Pastoral House
25 The Terrace
PO Box 2526, Wellington
Tel: 0800 CLIMATE (254 628)
Contact this person
