SENSITIVITY ANALYSIS AND VALIDATION OF A BIOGENIC
EMISSION MODEL:
BIOEMI 2.0
GRADUATION PROJECT FOR MONICA ROLLAN GALINDO
(November, 1998)
NATURAL POLLUTANS UNDER STUDY:
-
Isoprene
-
Monoterpenes
-
NOx = NO + NO2
BIOGENIC EMISSION METHODOLOGIES
BIOEMI 1.0:
ISOPRENE ALGORITHM (Guenther et
al., 1993):
Shows the influence of the PAR and the temperature in the isoprene emissions:

where CL is the light dependency:
and CT is the temperature dependency:
MONOTERPENES ALGORITHM (Guenther et
al., 1993)
Shows the influence of the temperature in these emissions:

where,
where,
T is the ambient temperature in ºK
b is 0,09
TS is 303ºK, the medition standar temperature
NOx
ALGORITHM (Guenther et al., 1993)
The NOx emission depends on the temperature:

where C y TS takes their values depending on the land use:
BIOEMI 2.0:
ISOPRENE ALGORITHM (Guenther et
al., 1993):
It is the same algorithm implemented in BIOEMI 1.0 and shown before.
MONOTERPENES ALGORITHM (Steinbrecher
et al., 1997):
Monoterpene emissions depends not only of the temperature but de PAR.
Adds a new factor to the Guenther algorithm that shows this dependency.

where,

NOx ALGORITHM (Yienger et al., 1995):
Adds new parameters:
-
Canopy reduction factor
-
Precipitation factor ("pulse")
More sensitivity with the land use type (more emission factors)
where,
¦ w/d(Tsoil,
Aw/d(biomass)) is a fuction that can be constant, lineal or
exponential
Aw/d(biomass) is the biomass coeficient
Tsuelo is the soil temperature
P(precipitation) is a factor used to adjust the emissions when the soil
"pulses"
CR(LAI,SAI) is the canopy reduction factor, due to the reabsorbation
by the plants of part of the emissions
w/d wet or dry
LAND USE MAPS
We use three different maps of the same area, the Comunity of Madrid,
for studing the sensibility of the modell to the land use map.
For studing the influence of the resolution of the map in the data results,
we have also simplified the INFOCARTO map (30 m.), and two new maps have
been obtained from the original, one of 250 m. and another of 1400 m.
REMO MAP
SOURCE: Environmental Software and Modelling Group. LANDSAT-V
1984 image
RESOLUTION: 250 m.
These are the land use types observed in this map:
-
Decidious forest
-
Perennial forest
-
Mixed forest
-
Olive
-
Garden
-
Bush
-
Pasture
-
Vineyard
-
Fruit
-
No vegetation
-
Rice
-
Water
-
Urban
-
Suburban
CORINE MAP
SOURCE: Fraunhofer Institut fuer Atmosphaerische Umwltforschung
(IFU), Garmisch - Partenkirchen in TFS-Proyect A6 (Dr. Gerhard Smiatek)
RESOLUTION: 1500 m
The land use types of this map are:
-
Decidious forest
-
Perennial forest
-
Mixed forest
-
Marsh or Wetland
-
Agriculture
-
Pasture
-
No vegetation
-
Tundra
-
Ice
-
Tropical/subtropical forest
-
Scrub/herbaceous vegetation
-
Water
-
Urban
INFOCARTO MAP
SOURCE: INFOCARTO, S,A. LANDSAT-V, IRS-1C, 1997
RESOLUTION: 30 m.
The land use types shown in this map are:
-
Decidious forest
-
Perennial forest
-
Mixed forest
-
Wet agriculture
-
Dry Agriculture
-
Dry Agriculture
-
Pasture
-
No vegetation
-
Erial
-
Bush
-
Water
-
Urban
-
Mosaic of natural vegetation and agriculture
RESULTS
COMPARISON BETWEEN BIOEMI 1.0
AND BIOEMI 2.0 ALGORITHMS FOR NOx AND MONOTERPENES:
NOx:
Seasonal Emission Patterns:
WINTER
SUMMER
SUMMER WITH MAXIMUM PULSE
MONOTERPENES
Seasonal Emission Patterns:
WINTER
SUMMER
RESULTS COMPARISON USING DIFFERENT
MAPS:
EACH MAP SHOWS DIFFERENT RESULTS FOR ALL THE BIOGENIC POLLUTANS STUDIED,
BECAUSE OF THE PERCENT OF EACH LAND USE TYPE THEY HAVE
-
REMO MAP: Biggest isoprene emisor
-
INFOCARTO MAP: Biggest monoterpenes and NOx emisor
-
CORINE MAP: Second monoterpenes and NOx emisor
LAND USE MAP RESOLUTION SENSITIVITY
- Little diferences between 30 m and 250 m results
- Big diferences between 30 m and 250 m results
- Diferences increases with PAR and temperature
WINTER, 30 m AND 1400 m MAPS
WINTER, 30 m AND 250 m MAPS
SUMMER, 30 m AND 1400 m MAPS
SUMMER, 30 m AND 250 m MAPS
RESULTS VALIDATION USING GEIA
INVENTORIES
-
ISOPRENE: The emission pattern adjusts, but it is needed a correction
factor (10 when using CORINE map, 7 with INFOCARTO map and 4,5 with REMO
map)
BIOEMI 2.0, REMO MAP
-
MONOTERPENES: The emission pattern adjusts. The best results are
obtained using the INFOCARTO map.
BIOEMI 2.0, INFOCARTO MAP.
BIOEMI 1.0, REMO MAP
-
NOx: The emission pattern adjusts. The best results are obtained
using the INFOCARTO map.
BIOEMI 2.0, INFOCARTO
MAP.
BIOEMI 1.0, REMO MAP
WHAT IS GEIA?
GEIA (Global Emission Inventory Activity)
GEIA is a component of the International Global Atmospheric Chemistry
(IGAC) Project.
1 x 1 degree resolution
Data:
-
ISOPRENE: 1990 monthly inventary, mg C/ m2 month
-
MONOTERPENES: 1990 monthly inventary, mg C/ m2 month
-
NOx: 1995 monthly inventary, mMoles N/ m2 month
http://blueskies.sprl.umich.edu/geia
POLLUTANTS SURFACE DISTRIBUTION:
Monoterpenes and isoprene are emitted in the norh of the study area,
where we can find the forest land use type, while NOx are emitted in the
south, where the agriculture land is distributed:
CONCLUSIONS
PRINCIPAL BIOGENIC POLLUTANTS
-
MONOTERPENES: They are the most emitted biogenic pollutans in the
area, during all the year.
-
ISOPRENE: It is the second emitted pollutant in summer and spring,
due to its sensitivity to these parameters.
-
NOx: The second emitted pollutants in winter and autumn.
BIOEMI 2.0 IMPROVEMENTS
-
MONOTERPENES: Same results in winter and night with BIOEMI 1.0,
but there is a big increase in emissions when PAR increases (summer and
day) with the new metodology.
-
NOx: Big increase all the year, specially when "pulsing". Sensibility
to more parameters, like precipitation, canopy, etc.
RESOLUTION SENSITIVITY
-
No significant differences when using a 250 m resolution map instead of
a 30 m.
-
Big differences when using 1400 resolution map.
These differences depends on the land uses that desapear when simplifing
the map (land uses with little percent of apareance or very disperse tends
to desapear with higher resolution).
LAND USE MAP SUGESTED RESOLUTION FOR
THE STUDY AREA:
-
CELL SIZE < 250 m => NO SIGNIFICATIVE CHANGES
-
CELL SIZE > 1400 m => SIGNIFICATIVE CHANGES IN ISOPRENE AND MONOTERPENES
EMISSIONS.
-
CELL SIZE > 1400 m => GREATE CHANGES IN NOx EMISSIONS.
-
BIG RESOLUTION => GREAT USE OF RESOURCES (TIME AND SPACE).
Because of these results, we estimate that the optimal cellsize for the
study area is 250 m, that gives quite good results, and expends much less
resources.
LAND USE MAP SENSITIVITY
-
Less important than the algorithm impact
-
REMO is the highest isoprene emisor because it has more deciduous forest
-
INFOCARTO is the highest monoterpenes emisor because it has more perennial
forest
-
NOx: In NOx emissions there are more uses than affect the results (all
types of agriculture, pasture, etc), being INFOCARTO map the highest emisor,
but with little differences with CORINE map.
MODEL VALIDATION
-
All the emission patterns adjust to the GEIA inventories.
-
ISOPRENO needs a correction factor (10 for CORINE map, 7 for INFOCARTO
map, 4’5 for REMO map)
-
INFOCARTO map obtains the best results with monoterpenes and Nox (it is
the highest emitter)
-
REMO map is the best with isoprene (it is the higest emitter)
-
New algorithms adjust better to GEIA data than the algorithms used before.
If you wish to see a Full Version of the GENEMIS Workshop (1998) where
this work was presented, please click here