ent snow, Built-up area
igh productivity, High
OW productivity, Non-
1 for Coniferous forest,
sted peat bog.
0: Vigorous vegetation,
getation, Bare rock and
gures
uting numbers for 430
10le country.
done automatically by
es a PSQL-command
statistical level starts an
ueries to different tables
acreage figures for the
itten to an XML -file.
s about 25 seconds, 65-
23 hours for the whole
on Skog og landskap's
1 to the acreage figures,
> statistics, the involved
read and nice to look at,
sures a nice layout of the
save work, one CSS-file
sated. CSS defines how
tyles saved in this CSS-
ce and layout of all the
yy editing this single file
car, and acreage figures
jle. People may look at,
jurce statistics sheets as
ce statistics for Norway:
>r/2010/Norge arstat 20
verage, but is inhomoge-
tion, and update frequen-
pecific importance more
luct sample-based moni-
r monitoring of the agri-
e (Dramstad et al, 2003)
ape phenomena based on
ites (1x1 km squares) by
al photos. Each square 15
record changes. Four 1n
ndscape spatial structure,
nd accessibility.
The specific monitoring of changes in the agricultural landscape
combined with the long-term changes captured by the full cov-
erage area resource statistics provides information that makes it
possible to analyse trends and drivers in the changes.
Further, the processing of 3Q-data was improved based on the
successful use of PostGIS for ARSTAT. The landscape analysis
application was a typical desktop GIS production flow, includ-
ing several tools, format conversions etc. We decided to fully
exploit the GIS-functionality integrated in the DBMS while
“translating” the application to PostGIS. A significant reduction
in complexity and processing time was achieved.
The GIS functions in PostGIS cover the Simple feature specifi-
cation (ISO, 2004) and more. This makes it possible to analyse
spatial information and compute and store the entire set of indi-
cators using relatively simple SQL-queries.
5. CHALLENGES
The statistics are based on ARS and AR50. ARS has the most
detailed classification. AR5O is based on generalized ARS-data,
the topographic dataset N50 and interpretation of satellite
images in mountainous areas. ARS is yearly updated, while
AR50 is revised every second year. The land resource statistics
is accordingly produced every second year. Changes in the ARS
dataset may therefore not be published in these figures before
two years after the changes actually took place.
A copy of ARS is frozen as a “year version” of the dataset in
January, e.g. year version 2011 is the database version available
at the beginning of 2012. During 2012, ARS will be updated.
These changes will not be included before the year version 2012
produced January 2013.
The changes in ARS occur at different time in different
municipalities and landscape types. Thus, the real changes in
the landscape are better presented using longer intervals than
one year. Although we consider the production line to be fairly
efficient, there are small changes from year to year on a national
scale. In the future, in order to better highlight significant
changes, the statistics will be issued every third year.
Total acreage numbers from ARSTAT may differ a little bit
from the official total numbers given by NMA due to different
Map projection, data source and generalization. NMA has used
Lambert azimuthal equal area projection (ETRS89-LAEA) with
azimuthal origin in the centre of gravity of Norway (E12, N64)
for their dataset with administrative borders and acreage figures
for all municipalities in Norway. AR5 and ARS0, on the other
hand, are given in ETRS89-UTM33.
An additional challenge for this year's version was the new high
quality version of Norway's coastline published by the NMA in
October 2011. We decided to use this new shoreline in order to
Correct our data in some remote arcas. A side effect was that
this introduced a lot of slivers where the old coastline was quite
OK. Norway has the world's second longest coastline (after
Canada). This overlay process allowed us to try out several
Citing functions in the PostGIS toolkit.
6. CONCLUSIONS
The open software solution is reliable, stable and fast. It is quite
Cas) to use and to adapt to PostGIS if you know a little bit of
SQL and are familiar with standard GIS functions.
Publishing statistics on national to local level based on one data
source is in many respects efficient. While data capture
(through remote sensing or field work) is still costly, the use of
efficient GIS tools offers flexible data processing to a low cost.
However, publishing statistics based on fairly complicated and
inhomogeneous data also poses quite a bit of challenges, espe-
cially communicating the information to the users. The need for
high quality data capture of the real landscape should not be
forgotten.
7. REFERENCES
References from Other Literature:
Heggem, E., Strand, G.-H., 2010. CORINE LAND COVER
2000, The Norwegian CLC2000 project. Skog og landskap.
Report 10/2010.
http;//www.skogoglandskap.no/filearch ive/Rapport 10 10 cori
ne land cover 2000-1..pdf
Aune-Lundberg, L., Strand, G.-H., 2010. CORINE LAND
COVER 2006, The Norwegian CLC2006 project. Skog og
landskap. Report 11/2010.
http//www.skogoglandskap.no/filcarchive/Rapport 11 10 cori
ne land cover 2006-1..pdf
Bjerdal, L, Bjerkelo, K., 2006. ARS klassifikasjonssystem.
Klassifikasjon av arealressurser. Skog og landskap. Manual
01/2006.
http//www.skogoglandskap.no/publikasjon/1170254097.17
Statens kartverk 2012a. N50 Kartdata. Brochure.
http://www statkart.no/?module=Articles;action=ArticleFolder.
publicOpenFolder;ID=5674
Skog og landskap, 2011. Arealressurskart ARS, AR50, AR250,
CLC. Skog og landskap. Brochure 1/2011.
http://www.skogoglandskap.no/filearchive/arcalressurskart bro
siyre.pdf
ISO, 2004. Geographic information -- Simple feature access --
Part 2: SQL option. ISO 19125-2:2004.
http://www.iso.org/iso/iso. catalogue/catalogue te/catalogue de
tail.htm?csnumber-40115
References from Journals:
Gjertsen, A. K., Angeloff, M., Strand, G.-H.,. 2011.
Arealressurskartover fjellomrádene. Kart og plan, Aas, no.
71/1, pp. 45-51.
http://www .skogoglandskap.no/publikasjon/arcalressurs kart ov
er fjellomradene
Dramstad, W. E., Fjellstad, W. J., Strand, G.-H., Mathiesen, H.
F., Engan, G., Stokland, J. N., 2002. Development and imple-
mentation of the Norwegian monitoring programme for agricul-
tural landscapes. Journal of Environmental management, 64,
pp. 49-63.
http//www.skogoglandskap.no/publikasjon/11877041806.1
References from websites:
Tenge, I. M., 2011. Det blir mindre kvalitetsjord. Forskning.no,
September 8 2011, 1 page.
http:/www.skogoglandskap.no/filcarchive/det blir mindre kva
litetsjord.pdf
