Quality of future N 2 O emission estimates: how might we improve? Keith Smith School of GeoSciences University of Edinburgh, UK.

Quality of future N2O emission estimates: how might we

improve?

Keith Smith

School of GeoSciencesUniversity of Edinburgh, UK

Possible areas of focus

• 1. Reviewing/revising default EFs for use when better data not available

• 2. Linked with (1), restructuring of some aspects of the methodology

• 3. Doing what we can to follow the Good Practice approach of using country- or regional-EFs where data can be obtained (“Tier 2”)– Not just using the default as the easy way out.– Including pooling information for neighbouring

countries/ countries with similar climates/soils/farming systems.

Default EFs

• Currently 1.25% of all N applied is assumed to be emitted as N2O.

• As later slides indicate, much evidence that this is not so, and disaggregation of various sorts is desirable, e.g:– By crop– By climatic zone– By the type of N applied

Some effects of the type of N

• Evidence is accumulating that mineral N fertilisers containing nitrate (e.g. ammonium nitrate, or CAN) added to wet soils, e.g. wet grassland in early spring, can give rise to much bigger emissions than those that are ever reported from urea or ammonium-N fertilisers.

• Manure spread evenly over fields yields less N2O than mineral N forms.

• In contrast, urine spots from grazing animals can give higher % emissions, influenced by the state of the soil.

Possible restructuring of the present methodology

• The calculation of N2O emission from grazed land could be dealt with under “Direct emissions from soils”, where it more logically belongs.

• Specific EFs could be adopted, if a consensus can be found, for (a) the manure N deposited in situ, taking into account the state of the soil under the grazing regime; and (b) the manure from animal housing etc spread on the fields (see previous slide).

Possible restructuring of the present methodology (cont.)

• Consideration might be given to replacing the complicated separate calculations concerning N input from legumes and from other crop residues, with a combined system for grain legumes and all non-leguminous crops – Estimating the N remaining in the field from the above-ground

parts, but crucially (and as a new development), including the root N.

• This would have to be complemented by separate calculations for forage legumes such as alfalfa, where all the shoot is carried off (see Robertson et al., Science, 2000), and clover-grass mixtures .– The role of rotational renewal of grass/clover leys by ploughing

and reseeding every few years also needs attention.

Comparison of some UK research findings with general assumptions used by Intergovernmental

Panel on Climate Change (IPCC) in existing methodology

IPCC:

“Default” emission factor EF):

1.25% of N applied to land emitted as N2O

-- same for all N inputs (fertilisers, manures, organic composts etc)

-- same for all crops

-- not dependent on soil properties, whether innate ones such as clay content, or transient ones such as wetness and temperature.

Results from work in UK:

EF varies with soil conditions, by up to 2 orders of magnitude (<0.1% to approaching 10%)

-- main factors are wetness, temperature, supply of nitrate

-- EF varies between fert. types

-- EF for cereal crops < IPCC value

-- but EF for some other crops and intensive grassland generally > IPCC value.

(see next 2 slides)

Contrasting approx. log- normal distributions of annual EFs:

top: all crops;centre: cut grassland;bottom: arable crops.

(a)

(b)

(c)

0.1-0.2 0.2-0.4 0.4-0.8 0.8-1.6 1.6-3.2 3.2-6.4 6.4-12.8

Fre

quen

cy

0

2

4

6

8

All crops

Emission factor (%)

0.1-0.2 0.2-0.4 0.4-0.8 0.8-1.6 1.6-3.2 3.2-6.4 6.4-12.8

0

2

4

6

8

0.1-0.2 0.2-0.4 0.4-0.8 0.8-1.6 1.6-3.2 3.2-6.4 6.4-12.8

0

2

4

6

8

10

12

Cereals/OSR

Veg. crops

Grass

All crops

Here, results of work in the UK (Dobbie et al, 1999; Dobbie & Smith, 2003) show values peaking well above the default EF, for grass, but well below for cereal crops.

Inter-seasonal variation in annual emissions of N2O from cut grassland, site near Edinburgh, Scotland, fertilised 3 times per season with ammonium nitrate:

-- consequence of variations in amount of rainfall around the times of fertilisation

Some other European studies strongly indicate EFs with average values greater than the IPCC default value of 1.25%, even though Bouwman et al. (2002) have produced a new global average value of 0.9%.

In the figure below, the regression line indicates a mean EF of 2.5%: twice the IPCC default value.

Relationship between total N input (by synthetic fertiliser, cattle waste and crop residues) and annual N2O emission measured on differently managed fields at Scheyern research station, South Germany. The regression excludes one data point (■). (From Flessa et al., 2002).

N2O emissions (continuous curve), soil nitrate contents,

0-30 cm (●), N fertiliser inputs, crops grown in the

rotation and N2O-N emissions, in a field

experiment near Braunschweig, Lower

Saxony, Germany (from Kaiser et al., 1998).

Peaks following winter freeze-

thaw events

Freeze-thaw events: large peaks in winter, can contribute 50% of annual emissions, but no allowance for strong continental winters in the methodology.

Recent work in Japan (Koga et al., 2004) gives very similar results to European experience.

Relationship between the ratio of dry matter to N content of incorporated crop residues and N2O emissions during winter, in a field experiment near

Braunschweig, Lower Saxony, Germany (from Kaiser et al., 1998).

The C:N ratio of crop residues appears to be a key variable in determining the amount of N2O produced during winter

-- could this also be allowed for, or is this too complicated??