International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B-YF. Istanbul 2004
author of this work to continue Bac's investigations, with the aid
of modern geoinformation methods.
The aim of this project was to evaluate quantitatively and qualita-
tively the soil erosion in Dublany from 1922 to 1994. Usefulness
of applied photogrammetric, cartographic and geoinformation
methods was also considered.
The research was expected to provide answers for two main
questions:
1. How much the ground level has been changed during the time
span of 72 years?
2. How does the extent and spatial distribution of this process
look like?
2. DATA PROCESSING & ANALYSIS
The measurement was based on the comparision of the two di-
gital elevation models representing two ground levels: 1922 and
1994. The ground level change was expected to c.a. 1.5 m and
therfore the measurements had to be carried out with possibly
high accuracy.
Figure 3. Plot image tone change caused by different
superficial soil humus content
In the first step, the strip of 3 B&W panchromatic aerial photo-
graphs at scale 1:10 000 was scanned (with 24pm scanner aper-
ture) and then photogrammetrically processed. After field GPS
measurements, the strip was adjusted with accepted accuracy
(RMSE ca. 0.4672 px). DTM for 1994 was produced also with
the aid of photogrammetric methods (automatic generation based
on image correlation with manual editing). In the next step deta-
iled ortophoto (0.3m ground resolution) was generated. and then
used to calibrate large scale (1:2 880) elevation map from 1922.
The contour lines with 1m interval were digitised and 1922 DTM
was interpolated. After standardization of the type and geometry
(raster with 1m spatial resolution), both models were substracted
(DTM1994 — DTM1922) and the resulting “diference” map was
generated (pixel value represents the ground level change). To
complete the analysis and facilitate comparison of the results
several series of terrain profiles were generated. These profiles
were spatially compatible with those published in Stanislaw
Bac's article in 1928.
3. CONCLUSIONS
The results of the study led to the following conclusions:
1. Soil erosion processes go on with the same intensity as they
went in the beginning of the 20-th century.
2. With the influence o
factors some hill tops were lowered and the valley bottoms were
f
f the atmospherical and anthropological
Figure 4. Plots with high erosion intensity has high ground
surface change factor (see legend on poster)
raised. Amplitude of this change amounts up to 4m (Fig. 4).
3. General tendency of ground flattening can be clearly seen.
Rain water moves soil particles down the slope and then deposi-
ted material is further moved away from the basin.
4. Erodibility of the Dublany loess soils is very high and it is
furthermore strengthen by agrotechnical measures (tillage). Plots
on still cultivated slopes are the most severely eroded.
5. Unproper agricultural practices (e.g. up and down tillage) and
lack of applied conservation methods in the past decades might
have had an influence to speed up the soil degradation.
6. Significant change of the landcover and land-use occured
during the investigated period. After 72 years the percentage
of built-up area considerably increased. These changes seem to
have a serious impact on the ground surface level.
In these circumstances, it might be worthy to note that precise
determination of all factors responsible for ground displace-
ment over particular locations is simply impossible due to the
aforementiond antropogenic changes of the landcover and the
land-use.
4. COMPLEMENTARY ANALYSIS
[n the second part of the project the quantitative results of photo-
grammetric measurements were compared with the results based
on the one of the most popular and widely used soil loss models
— USLE (Universal Soil Loss Equation), which was applied to
the same area. This analysis led the author to verify the effec-
tiveness of the USLE assessment of the average soil erosion.
Data model has been set up on the basis of the precipitation
data, topographic information (slope lenght and inclination) , soil
erodibility maps, land cover and landuse maps. The model was
previously adapted and validated in the south-eastern Poland in
'80 and '90. AIL information layers available for Dublany test-
124
Int
Si
th
c
gi
