Agriculture Reference
In-Depth Information
8.1.4 ESTIMATING EMISSIONS AT THE REGIONAL SCALE
The uncertainty in estimates of the global CH 4 emission from rice remains large:
the mean estimate for the 1990s in Table 8.1 is 60
30 Tg year 1 , equivalent to a
range of 5 to 15% of the total emission. This reflects the diversity of conditions in
which rice is grown and the large effects of management. To date, most estimates
have involved a down-scaling approach in which knowledge and understanding
at the field or local scale are used to extrapolate to the regional scale and above;
this approach is limited by the availability of reliable data at the required scale.
An alternative approach is to work in the opposite direction, down-scaling by
interpolation from measurements of overall terrestrial emissions made in global
air sampling networks (Heimann and Kaminski, 1999). Further improvements in
estimates are likely to come from a meeting in the middle of these approaches.
I here outline their pros and cons.
±
Up-scaling using Mechanistic Models and GIS
Knox et al . (2000) and Matthews et al . (2000b) have coupled a field-scale model
of CH 4 emissions from rice to GIS systems, and used available regional data on
weather, soil, agronomic management and other variables to make regional-scale
estimates of emissions. The model is based on the approach described earlier.
The extrapolation is based on the following framework.
Basic polygons for attributing vector datasets were derived from a digitized
map of national and provincial or state boundaries for China (31 polygons), India
(31), Indonesia (26), Philippines (80) and Thailand (73). At least one polygon
was defined for each province or state. Data on crop production and cropped
area under the four rice ecosystems were obtained for each polygon using the
map of Huke and Huke (1997), mainly from 1990. Weather data in individual
polygons were obtained from the nearest of 46 weather stations within the appro-
priate agroecological zone. Soil data were obtained from FAO-DSMW soil units
(1:5 000 000) with supplementary data for individual soil units in top and sub-
soils: pH, organic C, Fe content, texture and available water capacity. Each soil
property was given a weighted mean value for the polygon based on the distri-
bution of FAO soil units. Locally recommended crop management was assumed.
Due to lack of data, no allowance was made for differences in applications of
organic manures, and this will probably have caused underestimates in emissions.
The resulting global estimates are in the range 10-25 Tg CH 4 year 1 .
This approach is of course limited by the availability of reliable data and the
resolution of the data. An inherent problem in the 'up-scaling' process is the
interaction between variance in input parameters and non-linearity in models.
This may produce chaotic behaviour. van Bodegom et al . (2002) discuss this
in relation to CH 4 emission from rice. The point at which input data are aver-
aged before making model runs may also be limited by the available computing
Search WWH ::




Custom Search