Basic Approach to Mapping Different Sources, and the

Report
Basic Approach to Mapping Different Sources, and the
Sources of Spatial Datasets
John van Aardenne
jva@eea-europa.eu
European Environment Agency
Outline
1.Introduction
2.Reporting requirements
3. Wake up quiz
4. Post-processing of national inventory data
5. Gridding: concepts and datasets
6. Gridding: how does it work in practice
7. Visualization
European Environment Agency
1. Introduction
This presentation is aimed at providing a basic overview for those new
or relatively new to emission gridding.
Disclaimer:
Your presenter is not a GIS expert, nor a programmer, but with
common sense, database knowledge and nice GIS colleagues managed
to work on spatially resolved emission inventories starting from
simple scaling emissions with population (Moguntia model Nox
emissions), to EDGAR-HYDE AP and GHG emissions (1x1 degree),
Historical AP and GHG emissions for IPCC AR5 (0.5 degree) and
EDGARv4 (0.1 degree).
So there is hope.....
European Environment Agency
2. Reporting requirements: EMEP grid
Extendedd 50 x 50 km2 grid
Number of grid cells: ~21000
Size of grid cell
at 40°N (Italy): 40x40 km2
at 60°N (Scandinavia): 50x50 km2
European Environment Agency
2. Reporting requirements: EMEP grid
European Environment Agency
2. Reporting requirements: sectors
A. Public power
B. Industrial comb. plants
C. Small combustion plants
D. Industrial process
E. Fugitive emissions
F. Solvents
G. Road – rail
H. Shipping
I. Off road mobile
J. Civil aviation (domestic lto)
K. Civil aviation (domest cruise)
L. Other waste displacement
M. Wastewater
N. Waste incineration
O. Agricultural livestock
P. Agriculture (other)
Q. Agricultural wastes
R. Other
S. Natural
T. International aviation (cruise)
z. Memo
European Environment Agency
wake up quiz
Imagine the emep grid........
European Environment Agency
3a. The following grid cells represent……
1. Hungary
2. Austria
3. Latvia
European Environment Agency
3b. The following grid cells represent……
1. Malta
2. Liechtenstein
3. Luxembourg
European Environment Agency
3c. The following grid cells represent……
1. Belgium
2. The Netherlands
3. Turkey
European Environment Agency
4. Post-processing of emission inventory data: Emission inventories as
annual total by sector are not sufficient to allow for atmospheric chemistry
modeling
The EMEP unified model has 20 height
layers (www.emep.int)
Seinfeld, J.H. and Pandis, S.N., Atmospheric chemistry and
physics: from air pollution to climate change, Wiley and
Sons, New York, 969-971, 1998.
European Environment Agency
4. Post-processing of emission inventory data: 3 activities
are needed.
Horizontal allocation: assigning emissions to their proper grid
cell using gridded data on spatial surrogates with known geographic
distributions
Vertical allocation: assigning emissions to their proper layer in
the atmosphere. Often static vertical distribution factors are
applied to the emissions of each sector or all emissions are put into
the lowest layer.
Temporal allocation: representing emissions variation over time
(closure of facilities for maintenance, rush hour, weekends,
public holidays)
Source: US EPA emissions modeling clearinghouse, Bieser et al., 2011.
European Environment Agency
5. Gridding: conceptual
Point source: an emission source at a known
location such as an industrial plant or a power
station. (could be an LPS, or not, depending on
threshold)
The grid cells representing
the geographic domain for
which you have emissions
data.
Area source: sources that are too numerous or
small to be individually identified as point sources or
from which emissions arise over a large area
(agricultural fields, residential areas, forests)
Line source: source that exhibits a line type of
geography, e.g. a road, railway, pipeline or shipping
lane
The sum of all different types of emissions in your domain
European Environment Agency
5. Gridding: in formula
European Environment Agency
5. Gridding: conceptual
European Environment Agency
5. Gridding: the “trick” is to find spatial proxies to allocate emissions to a
specific grid. You will see several examples today, here results from a
recent publication. Bieser et al. 2011 SMOKE for EUROPE
Sector
Proxy 1
Proxy 2
Proxy 3
Proxy 4
Combustion in
energy and
transformation
industries
E-PRTR
CORINE land cover
(CLC; commercial
and industrial
units)
Global land
cover
database
(urban area)
Population
(GWPv3)
Non-industrial
combustion
Population
(GWPv3)
-
-
-
Road transport
TREMOVE
Open street maps
and Digital chart
of the world
(motorways,
roads)
CORINE land
cover (urban
area)
Global land cover
(urban area)
Agriculture
CLC
(agricultural
areas, pasture)
GLC (agricultural
areas)
EUROSTAT
(animal
stocks,
employees
European Environment Agency
agriculture)
5. Gridding: recently released high resolution dataset (100m)
Gallego F.J., 2010, A population density grid of the European
Union,Population and Environment. 31: 460-473
.
European Environment Agency
5. Gridding: population density with coverage also for non-EU countries and
split in urban and rural can be found at:
http://sedac.ciesin.columbia.edu/gpw/
NationalBoundaries and GPWv3 2005 Pop Density
European Environment Agency
5. Gridding: example CORINE land cover by NUTS unit
(http://dataservice.eea.europa.eu/PivotApp/pivot.aspx?pivotid=501)
European Environment Agency
6. Gridding: how does it work in practice
1.Define “your grid” cells
2. Define the different spatial allocation proxies
3.Calculate the fraction of spatial proxy in each grid cell
4. Separate point source emission information from national
total and allocate remaining emissions by sector with spatial
proxy
5. Saving time: combine sources with same spatial proxy
European Environment Agency
6.1 Define “your” grid cells. An example from the world on 0.5 grid
showing country boundaries based on GWP data
European Environment Agency
6.1 Define “your” grid cells. What you are seeing is this file table
with grid locations (lon-lat) and definition of countries.
This plot is nothing more than a x and y coordinate to identify the
grid cell and the corresponding value telling which country is found
in the grid cell.
European Environment Agency
6.2 Define the different spatial allocation proxies
See chapter 7 of the Guidebook, Appendix A Sectoral guidance for spatial
emissions distribution.
- apply those proxies that are associated with the emissions
- but also ensure to identify proxy data that would give strange results
(e.g. large fraction of wood combustion in London or Paris, when using
population as proxy)
European Environment Agency
6. 3 Calculate the fraction of spatial proxy in each grid cell
Grid=id
Emep Emep A
(i)
(j)
1
66
2
3
B
1
67
39 0.8
39 0.1
66
40 0.1
0
C
D
E
0
With for e.g. A: Population: B. Urban population, C. Road network
(%km, or using traffic count data)
Number of grid cells (EMEP domain):
Current grid: 21000
0.5x0.5: 25000
0.1x0.1: 624000
European Environment Agency
6. 3 Calculate the fraction of spatial proxy in each grid cell
Gridding: if spatial proxy data are available in the required grid
resolution you can start using excel on course resolution grids but
same principle can be build in a database environment
If the spatial dataset (e.g. population, traffic density) has to be build
from the original datasource, this part is of course less straightforward
(other presentation will confirm this).
For non-standard datasets and high resolution grids, you need GIS
support (software and staff)
For TFEIP, can standard datasets on proxies be made available if other
projects have already done the work (e.g. E-PRTR diffuse emissions,
FP research project, etc.)?
European Environment Agency
6. Gridding: how does it work in practice
1.Define “your grid” cells
2. Define the different spatial allocation proxies
3.Calculate the fraction of spatial proxy in each grid cell
4. Separate point source emission information from national
total and allocate remaining emissions by sector with spatial
proxy
5. Saving time: combine sources with same spatial proxy
European Environment Agency
7. Gridding visualization: you will see many examples in the
following presentations, ask what software they are using 
Nox emissions from surface fuel combustion (Dignon 1992), Image courtesy.
Schumann, U., A. Chlond, A. Ebel, B. Kärcher, H. Pak, H. Schlager, A. Schmitt, P. Wendling
(Eds.): Pollutants from air traffic - Results of atmospheric research 1992-1997. DLRMitteilung 97-04, 291 pp. DLR, Köln, Germany, 1997.
European Environment Agency
8. Do we have more certainty if we go for higher resolution emission
inventories? Some thoughts.....
We are getting access to
datasets with ever increasing
spatial resolution (e.g. 100
mtr population, exact location
of point sources),
Inventories: more work, nicer
pictures, more (un)certainty?
Seinfeld, J.H. and Pandis, S.N., Atmospheric chemistry and
physics: from air pollution to climate change, Wiley and
Sons, New York, 969-971, 1998.
Model results: better match
with observation, better
understanding of chemistry,
more aggregation of emissions
due to mismatch model
resolution?
European Environment Agency

similar documents