×

You are using an outdated browser that does not fully support the intranda viewer.
As a result, some pages may not be displayed correctly.

We recommend you use one of the following browsers:

Full text

Title
Technical Commission VII


CV pm fV
wu P
m-— VJ
tv

International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XXXIX-B7, 2012
XXII ISPRS Congress, 25 August — 01 September 2012, Melbourne, Australia
IDENTIFICATION OF LAND COVER IN THE PAST
USING INFRARED IMAGES AT PRESENT
V. Safát ?, V.Zdímal ^"
* GEODIS BRNO Ltd., Lazaretní 11a, 615 00 Brno, Czech Republic — vaclav.safar@geodis.cz
? Faculty of Agronomy, Mendel University in Brno, Zemédélská 1, 613 00 Brno, Czech Republic —
zdimal@mendelu.cz
Commission VII, WG VII/4
KEY WORDS: Remote Sensing, Land Cover, Land Use, History, Infrared, Manuscripts, Agriculture
ABSTRACT:
Czech landscape is an old residential area used by humans since ancient times. People have influenced it since their arrival and
various activities in different periods create landscape layers called a palimpsest. Land Cover of one location could have changed
several times. The most important reason is meandering and subsequent straightening of rivers, deforestation, relocation and change
in soil layers. These changes in the past affected the present management and it is important to identify them. A suitable tool for the
determination of different sites is remote sensing in the infrared spectrum, which monitors changes in the vegetation with the support
of archival materials. After identifying the different places you can search the archival materials, how the land cover looked in the
past. There have been used these archival materials: maps II. and III. military mapping, basic maps and other maps and historical
orthophotomap. Czech Republic has a national archive of aerial photographs with aerial photographs from the thirties of the last
century maintained by MGHO Dobruska. A comparative analysis of Land Cover shows the increases and decreases in agricultural
land, changes in communication line elements, forest losses and increases, comparing the legal and actual status of the forest
boundaries and their changes over time, changes in the built areas and links to the surrounding countryside. Land Cover of this study
was created primarily with a visual interpretation of each area with their vectorization and assigning attributes to these areas and then
comparing each of archival materials.
1. INTRODUCTION
The landscape is constantly changing. The causes of these
changes are both natural, but nowadays mostly human. In de-
veloped countries with dense settlement is great pressure on the
landscape and there are many people interested in its use. Czech
landscape belongs to old development areas settled by a man
from old ages. The landscape has been under human influence
since people first arrived on the scene, with individual eras
leaving their mark in the form of layers which may be read al-
most like a palimpsest. A single location may show evidence of
human activity from the different periods and natural changes.
The Land Cover of one location could change several times.
The most important reason is meandering and following
straightening of rivers, deforestation and soil movement. In the
past those changes influenced today's management and it is im-
portant to identify them. When we design a new Land Use is
important to know the Land Cover in the past. Knowledge of
the Land Cover in the past allows us to propose the appropriate
the Land Use in the present and prevent unsuitable use of land-
scape with limited usage. One of the tools used to determine the
different places are remote sensing in the near-infrared band
monitoring the vegetation changes. These changes show the
vegetation places with different histories of others. After identi-
fying the different sites it can be found in archival materials,
what was the Land Cover in the past.
Lipsky [1999] and Kubes [1996] describe detailed origin and
development of cultural landscape in the Czech Republic.
Czech landscape underwent a fundamental change in the period
of socialist collectivization of agricultural production since
1954. The area of agricultural land increased, meadows and
grassland were arable in the lower altitudes. Usage of plant pro-
*
Corresponding author.
tection chemicals and chemical fertilizers negatively affected
the biota of agricultural land, neighborhood land and biota wa-
ter streams and reservoirs. Insecticides kill many insects, heavy
metals get worse health status and reproductive abilities of birds
and mammals, plant species sensitive to nitrogen disappeared
and plants able to live in the new conditions appeared. Increas-
ing the area of agricultural land and the change of management
of the meadows and pastures have deteriorated aesthetic param-
eters of the countryside. In some areas there are occurred in-
creasing of the size of area of woody vegetation of agricultural
landscape in the period of socialist collectivization of agricul-
ture. That development was different in each territory of the re-
public, depending on whether it was a production area or mar-
ginal. Large-scale socialist farming practices in the country left
a number of small unusable areas with weeds and trees. The
current state of the landscape is not very different from the con-
dition of the landscape during periods of maximum application
of forms of socialist agriculture. Examples of simplification of
the landscape microstructure are shown in Figures 1. and 2.
Spectral characteristic of plants are the main consequences of
the different land use. The spectral behavior of vegetation
showed a significant increase in reflectance in the near infrared
spectrum. It is commonly stated that in the visible spectrum
about 20% of incident radiation is reflected by vegetation, while
in the near infrared spectrum about 60% is reflected. The differ-
ence in the amount of chlorophyll in leafs has a great influence
on the determination of LAL An influence of the amount of
chlorophyll on the LAI determination was studied by Habou-
dane et al. (2004). Convenience of used spectral bands is often
discussed question. Primary, standard red (RED, 630-690 nm)
and near infrared (NIR, 750-900 nm) bands are used. Zhang et
al. (2006) used these spectral bands and NDVI, respectively, for