Air Pollution as a Climate Forcing: A Workshop

Day 1 Presentations

Declining Estimates of Methane Emissions from Rice Agriculture

Hugo Denier van der Gon
TNO Environment, Energy, and Process Innovation (TNO-MEP), Apeldoorn, The Netherlands

You may download a MS PowerPoint version (2.8 MB) of this presentation.


Rice is one of the world's most important food crops. Wetland rice fields constitute one of the major anthropogenic CH4 sources, but the source strength is surrounded by a large uncertainty. Estimates of the rice source strength have varied over the last decades between 12 and 280 Tg/yr. By studying the estimates and the motivations behind them, some conclusions can be made.

  1. Averaging all the previous estimates does not yield the best estimate: some of the previous estimates are based on wrong assumptions or do not include all the relevant information presently available.
  2. The annual CH4 budget is a snapshot; the estimates should not be static but changing as a result of technological and socio-economic changes.
  3. There are no proper constraints on the rice CH4 source strength. Therefore, each estimate is as good as the next one (or the previous one) and it is impossible to "prove" that an estimate is wrong.
  4. Since the global CH4 budget is relatively well constrained there is hesitance among atmospheric scientists to accept changes in individual source strengths because a dramatic change in one source should be compensated by changes in other sources. "Single-source" scientists do not deliver such alternatives; this reduces the acceptance of new estimates, especially in combination with point 3.

The bottom-up estimate for the rice CH4 source strength in the 1990s. The reasons for differing estimates are (1) incomplete understanding of the rice ecosystem and its biophysical and socio-economic dynamics and (2) poor data on the spatial and temporal variability of factors controlling CH4 emission.

Table 1: Selected bottom-up estimates of the rice CH4 source strength for the 1990s
Reference CH4 release (Tg/yr) Method
Sass (1994) 25-54 Review of CH4 studies in different countries
Neue (1997) 32 Review of CH4 studies in different countries
Wassmann et al. (2000) 10-25 Extrapolated from UNDP-funded monitoring program in 5 major rice growing countries
Denier van der Gon et al. (2001) 32-37a IPCC good guidance guidelines combined with a spatial database of rice ecosystems (assuming use of organic amendments on 40% of the acreage)
adepending on choice of harvested rice area in China


Table 2: Methane Emission by rice ecosystem calculated from total carbon returned to the rice soil
Rice ecosystem Methane emission (Tg/yr) calculated using
Standard rice acreage for Chinaa Alternative rice acreage for Chinab
Irrigated rice24.829.3
Rainfed rice7.27.4
Floodprone rice0.590.59
Upland rice00
Total rice32.637.3
a Huke, R E. and E. H. Huke, IRRI, 1997.
b Verburg, P.H. and Y.Q. Chen, Ecosystems 3, 369-385, 2000

Alternative calculation of the rice CH4 source strength. An independent proxy method can be used to estimate the CH4 emission from rice fields.

Trends in CH4 emission from rice fields. Since the mid-1960s, remarkable growth in rice productivity has been achieved through the diffusion of modern rice varieties, widely known as the green revolution (Fig. 1) Simultaneously — and part of the green revolution — the use of fertilizers in Asian (rice) agriculture has increased tremendously. As a result, the primary basis of growth in Asian rice production shifted from crop area expansion to increases in yields. These developments have important consequences on the annual CH4 emission from rice fields. For some time it was believed that increasing rice production would result in increasing CH4 emissions. However, our studies indicate that CH4 emissions per ha cultivated land have not increased despite an increase in yield and may even have declined due to reduced use of organic amendments (Fig. 2.).

Time chart of rice adoption. See caption and text for more.

Figure 1: The adoption of modern high-yielding rice varieties in 4 Asian countries. The adoption depends on the suitability of the rice ecosystems in the countries. The figure illustrates the dramatic rate at which technological innovations can spread in agriculture. The type of variety grown and the inputs needed to achieve the yields have a strong influence on CH4 emission. Therefore, there are trends in CH4 emission over time but they differ by country

Maps of climate forcings. See text for more.

Figure 2: Methane emission per hectare over time in a country with high adoption of modern rice varieties (Indonesia). The emission rate is derived from the rice yield statistics but achieved with different variety characteristics. If the adoption of modern varieties is taken into account the CH4 emission per ha of harvested rice land remains stable at ~200 kg/ha The higher emission from traditional varieties in Fig 2. is not because traditional varieties emit more CH4 per plant but because more biomass is needed to obtain the present day yields. Since increased production is needed to feed the population, the use of modern varieties has mitigated CH4 emission from rice. (Denier van der Gon, Global Biogeochem. Cycles, 14, 61-72, 2000)

Constraints on the rice CH4 source strength. Constraining the global rice CH4 source using inverse modeling of the CH4 sources and sinks is not possible due to (1) a lack of measurement stations directly influenced by terrestrial sources in the Asian region, (2) the low temporal sampling frequency which is insufficient to derive reliable monthly means, (3) the low resolution of the global model used for the inversion and, (4) the high uncertainty of other Southeast Asian sources of methane.

On the mitigation of CH4 emissions from rice fields. To reduce CH4 emissions from rice fields, rice-growing countries must try to persuade farmers to adopt mitigation technologies while avoiding any unnecessary burdens for them. Several mitigation options have been identified e.g., (1) Intermittent drainage in irrigated systems, (2) Improved crop residue management through composting or mulching, (3) Direct seeding and (4) The use of sulfate-containing amendments. So far, for only the last option a cost assessment has been made (estimated costs at 5-10 US dollar per Mg CO2-equivalent). In this widely varying agricultural production system, mitigation options will have to be tailor-made local or regional strategies; no mitigation option is applicable across the board.

Conclusions. The estimates of the rice CH4 source strength have been declining from 110 Tg/yr (range 25-170) (IPCC, 1990) to 60 Tg/yr (range 20-100) (IPCC, 1997). A best guess of the rice paddy source strength in the late 90s based on recent bottom-up studies would be 30±15 Tg/yr. We can derive some confidence from the convergence of emission estimates based on (1) bottom-up calculations with activities and emission factors, (2) upscaling with summary models, and (3) proxy methods. However, we are not able to truly constrain rice emission estimates at the intermediate scale. The number of field studies (flux measurements at the plot scale) is more than sufficient, but intermediate scale (landscape, regional or national) emission measurements are essential to validate regional CH4 budgets. Such measurements are technically feasible, but relatively expensive.

Due to innovations in rice farming, such as the high-yielding modern varieties, and a shift from organic manures to fertilizers, yield increases have not resulted in increasing CH4 emissions per unit of harvested area. It is likely that in some areas CH4 emissions have substantially decreased over recent decades. Optimizing rice agriculture can reduce CH4 emissions by optimizing the amount of carbon going into grain production and reducing the amount of carbon available for CH4 production. It is of great interest to understand how future yield increases, which are necessary to feed the growing population of rice consumers, will be achieved and what the consequences for CH4 emission from rice fields will be.

The implications of a lower rice CH4 source are that (1) rice may be one of the sources responsible for the slow down of the atmospheric CH4 growth rate in the 1990s, (2) the potential to reduce the global CH4 burden by mitigating the rice source strength is limited and much smaller than previously thought, and (3) modeling of the atmospheric CH4 sources and sinks suggests that a significant part of the global CH4 sources must be located in the tropics — therefore, if rice is a smaller source than previously believed, do we need another source to be larger?

Workshop Homepage * Background
Summaries: Overview, Gases, Aerosols, Tech., Health, Agri./Eco.
Abstracts: Day 1, Day 2, Day 3, Day 4, Day 5 * Participants