5449 50 51
82 53
Fig. 5: General view of recultivated areas of the opencast mine Welzow-Süd (mosaic of aerial photographs taken by the
camera AFA-41, 08. 10. 1993), scale 1 : 25 000
and degree of homogeneity during their temporal
development. Furthermore it is of interest to determine
the planned and undesired moist and wet spots as well
as the soil moisture and substrate distribution for areas
still without vegetation cover.
The investigations on monitoring recultivation processes
have been carried out in the area of the opencast mine of
Welzow-Süd (frame 2 in Fig. 1). The annual coal output
of this mine amounts to about 26 Mio tonnes of lignite
requiring approximately 200 Mio tonnes of stripping
material to be removed. As a result dump areas of
several thousand hectares have emerged. Fig. 5 shows
part of the dump with recultivation zones of different land
use.
In order to investigate the feasible contribution of remote
sensing on monitoring recultivation processes remotely
sensed features based on multisensor data have been
derived (Fig. 6). For description of the recultivation pa-
rameters to be searched for, combinations of these
remotely sensed derived features including ancillary data
(expertises on dump soil, recultivation planning, meteoro-
logical data, field survey of land use classes) have been
applied and processed.
One precondition of process monitoring is the multitem-
poral approach. The investigations presented here were
carried out on available satellite imagery taken from 1991
to 1995, comprising six Landsat TM scenes, one SPOT-
XS scene and two airborne scanner data sets (MSU-M).
Parallel to the visible data airborne thermal scanner and
SAR sensors were used to create multisensor data sets
thus enabling simultaneous data processing and analysis
to be performed during the acquisition times.
To attain comparability with the aim of deriving trends the
data had to be normalized and eliminated from over-
laying disturbing effects. Such a normalization procedure
Indices of Vitality
Multispectral CA (NDVI, VM2) ^ Object.
Scanner Data |— Catego ;
(VIS,NIR, |. iro ey
SWIR) TON Moisture and TT
A Soil Index J
E uc
pur I A Bsoluto ( Vitality )
EA pu) S
Thermal Infrared | — Temperatures T.
Scanner Data
LS. | Differences of Sent
»J| Temperatures er
P RAE n nei
ne
= Se
Microwave Data |—^ = EN b z
(SAR, Imaging | omi aed [7 s =
Radiometer)
Texture Features | — à ( Moisture )
(variance, ..) | "| S. ~
p 7 =. NE d
| To formed / Substra-
Auxillary Data ENS AZ \ tum )
NH ud cw wu S A D d
Fig. 6: Scheme of relationships between remotely sensed
data and recultivation features (VM2: 'Spectral Curvature
Index’, after Weichelt et al., 1987)
772
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B7. Vienna 1996
NIPYVI
NDVI
Fig.
aree
has
date
conc
led 1
crep
well
re d
refer
fores
mine
som
for s
It ca
ted 4
whic
Furtt
vitali
in su
tural
times
for p
whicl
short
May/
(area
detec
then
the ai
Base
term
the d
ring €
recult
1995.
restec
