4 - Fires
Start date or starting event: full duration of project
Fire-Climate-Carbon cycle interactions on regional and continental scale
The overall objective of the Fire WP is to quantify the current
African fire-related GHG emissions and fire-vegetation feedbacks,
and their regional and interannual variation using an integrated
remote sensing and process-based modelling methodology.
4.1 Development of an improved and validated method to estimate C
emissions from forest and savanna fires, combining the three
discrete, multi-sensor EO-based approaches of active fire detection,
burned area mapping, and fire radiative energy assessment.
4.2 Derivation of key datasets of fire variability and an analysis
of influencing factors, including climatic oscillations (e.g.. El
Nino), population density, land cover type and meteorology.
4.3 Analysis of driving factors of the fire regime in different
parts of the study region to improve fire modelling through
quantification of coupling of interannual climate variability,
population density, land cover type and fire frequency and intensity
using a remote sensing data analysis approach.
4.4 Incorporation of processes identified in (4.1) - (4.3) into
an improved fire module of a DGVM model framework (LPJ-GUESS).
Regional and continental model estimates of total fire carbon
loss, post-fire carbon-dynamics and fire-climate-vegetation
|Description of work |
Task 4.1 Interactive climate-fire-vegetation effects.
Task 4.2 Remote sensing of fires and estimates of carbon losses.
4.1 A first full year (2004) of fire radiative power and energy data
for the whole of African continent, at source (~3 km) spatial resolution
and 15 minute temporal resolution, derived from SEVIRI and compared to
MODIS in the study regions; together with a sensitivity analysis of
the FRP retrievals to parameterisation errors and algorithmic assumptions
4.2 Study region, MODIS-derived FRP datasets for 2002 and 2003 (Month 18).
4.3 Burned area map of study region from remote sensing (Month 18).
4.4 Integrated methodology combining burned area, active fire detection
and FRE into a single, optimised biomass burning C emissions product
for Africa (Month 22-27).
4.5 Field-tested relationship between fire radiative energy and carbon
emission for different African fuel types and a range of fuel moistures,
enabling the above FRP datasets to be converted to carbon flux estimates
from identified fires. Validate the fire module of the LPJ-GUESS model
using new remote sensing and GIS data, as well as information on
population density (Month 22-27).
4.6 Process-understanding of regional and continental fire C losses in
Africa, their relation to climate, area burnt and frequency and causes
for interannual variation. Asessment of fire-climate-vegetation
interactions on boundaries, e.g., savanna-forest and savanna-desert
that can be considered particularly vulnerable to climate change and
human pressure (Month 22-30).
|Milestones and expected result|
4.1 WP kick-off workshop in Skukuza, relatively early on in the
project to discuss our various approaches, including a one or two-day
course or so at Wits afterwards (Month 4).
4.2 Algorithms for fire identification and characterisation; Improved
fire routine in LPJ-GUESS; Revised PFT in LPJ GUESS, accommodating
fire characteristics (Month 12).
4.3 Analysis of matching SEVIRI- and MODIS-derived FRP/FRE datasets
for 2004. Burned area map of study region over 3 years (e.g., 2002-2004)
and matching MODIS-derived FRP data; Analysis fire-human interactions
(e.g., population density, land use, use of controlled fire) and
improved assessment of human impacts on fire in LPJ-GUESS (Month 18).
4.4 Validation of fire products using ground and airborne data, and
comparison to model results; First continental LPJ-GUESS model runs;
2nd WP workshop in West Africa, again with training workshop, lectures
4.5 Estimation of carbon emission from fires, and its spatial and
temporal variation in study regions/years (Month 27).
4.6 Analysis of causal factors in the fire regime using GIS and
climate data, comparison with model causal analysis. Regional and
continental model runs to estimate the effects of fire on African
carbon balance (Month 30).
4.7 Completion of manuscripts, publications (Month 36).