INTEGRATION OF PROCEDURES AVAILABLE IN LOW COST IMAGE
PROCESSING AND G!!:OGRAPHICAL INFORMATION SYSTEMS.
An applica-tion on human settlement and landscape ecology in
the depression of EI-Fayoum (Egypt).
Th.K.Ghabour & M.Antrop
Laboratory for Regional Geography and Landscape
Science. State University of Ghent - Belgium
C: 0 m . V I I
-
I.JG. 7
1. Introduction
Agriculture in Egypt. as in many other countries in the arid
and semi-arid regions, is suffering from salinization of the
soils. Soil salinity is one of the most serious problems in
agriculture under irrigation. A decline of soil productivity
of 30% may occur and is attributed largely to difficulties
in oPtimi=in~ the cOlnbined irrigation and drainage system.
However, this is not a technical problem in the first place,
but a social one. The increase of the population during the
second half of this century,
caused also an increasing
demand of water. A more strictly policy of the water supply
was needed, causing important social tenses.
2.
The depression of EI-Fayouul
El-Fayoum ref~rs to the administrative governorate which is
located in a deep. near-circular depression in the limestone
plateau of the northern part of the Western Desert of Egypt
{fi~.l}.
It is situated south-west of Cairo at a distance of
about 60 km and nearly 35 km to the west of the Nile. The
depression is limited at the north-western side by a shallow
brackish lake.
called Birkst carun, with an area of about
200 km~ and forming the deepest part at about 45 m below the
present sea level. The lake is fed by water of the Nile
brought into the depression for irrigation. About 1800 km~
of the depression is filled with rich alluvial sediments and
more than 1350 km~ is cultivated. This area represents about
5.3% of the total cultivated area of Egypt.
EI-Fav0um Governorate is divided into 5 provinces
ElFav()um,
Ibshat,Jai,
It.sa,
5innuris
and Tamiya
I, fig. 2).
Although th~
latest population census dates from 1986. no
data are available yet.
The census of
1976 gives
1.14
million people living in the governorate. which was about 3%
0f the
total population of Egypt.
The estimation of the
actual population gives 1.44 million inhabitants. The rural
population accounts for 76% and is mainly distributed in 140
villages or
small towns.
Thus the population density
increased f'rom 370 inh./km~ in 1947 to an estimated value of
a,br:"ut 8 1)(1
inh.,/km;;;' in IS1e-5.
The population is not equally distributed allover the
depression. but shows a
concentration in the central part
and on the most fertile soils. Consequently, the population
pressure increases here most rapid Iv and most of the fertile
soils are affected in the first place.
Salini~ation occurs
almost generally in the depression and
reaches dramatic values. Detailed monitoring of the
VU-48
evolu~ion
and the
tasl:s.
of the population pressure, the irrigation system
process of salinization become more and more vital
3. The problem.s
Many problems do occur to realize this kind of
landscape ecological monitoring. Some of them are
global
- accurate and detailed demographic statistics are not
available since the census of 19 7 6;
- the detailed demographic data which are necessary for
making prognoses of the future growth,
refer only to rather
large administrative units and do not reflect the real
population distribution of the settlements;
- inventories of the soil conditions are very general and do
not reflect the rapid changes in the environment;
- collecting and integrating all necessary information for a
regional and local land assessment is very difficult.
Clearly, the use of satellite remote sensing and the use of
geographical information
systems
{GIS)
seem to offer
interesting
solutions.
Unfortunately,
no
financial
possibilities to develop such operational systems,
nor a
sufficient number of trained researchers are available.
Consequently, the approach should be very pragmatic, using
every tool which is available at this moment. The general
idea is to integrate any kind of land information on a PC
standard, because these are more or less widely available in
administration. The sources for the land information system
are all existing maps,
aerial photographs,
statistics as
well as remote sensing data from Landsat TH.
4. Population pressure and salinization
This contribution presents only as an example one of the
first attempts to relate these different sources of data on
a regional and local PC-based Land Information System. Basic
questions were
ll)
determine an estimation of the actual population for
each settlement site;
~2) make an inventory of the actual situation of the soil
conditions related to the process of salinization;
~3)
assessment of the human pressure on the land;
l4) test if any relation between population pressure and
salinization exist. and if any determine 'hazard or highr i8k zc'nes'.
VII ... 49
Q...~..........$.~..t.!:.1.~.m.~n t.;.,~., ..,. .g..n~;.L
.
. ,.P.Q.P.V..b,.Q,.!:.,;!'.Q,.H
The population
is mainly concentrated in about 140 villages
about 200
to 3000 inhabitants. Each district
has also a town or even a small city. The villages are
almost randomly distributed allover the depression. They
can be detected and delineated easily upon Landsat
TM
imagery \on band 3 black & white and false color composites
with bands 2.3.4). Visual interpretation proved to be rapid
and accurate because of the natural capability of the human
percePtion in dealing with complex contextual
information
and because most
localizations are approximately known.
Shape and size could
be determined and compared with the
aerial photographs dated 1956.
Figures 4
and 5 show the
image ot
El Minva
(about 13000 inhabitants) on the aerial
photograph and on the Landsat TM (band 3). The new extension
of the town can be seen clearly.
Thus,
a
complete and
updated settlement mapping could be achieved and which is
presented in figure 2 where the administrative boundaries
are added also. No changes occurred in the territories and
sites of
the settlements during the last 40 years,
only
their sizes changed.
rangin~
from
A population census on a communal level is only available
for 1947. From the census of 1976 only data on the district
level are generally available. Both have been use to make an
estimation of the population of each village in 1985, the
year of the Landsat TM registration. A logistic growth model
was used.
The carrying capacity was estimated using the
territorial size.
Determining the carrying capacity is not
easy because of lack of appropriate ground data. especially
in the past. Simulation of the logistic model was used to
find the best fit
of the predicted population with the
general sensus data. This resulted in a rural
population
density of about 800
inhabitants/km~. Such
an approach is
only valid for rural settlements in which is initially
physical isotropic.
The results are strongly biased for
cities and
large territories containing heterogeneous land
qualities. Another estimation of the carrying capacity can
be based upon the village sizes. The correlation between the
population in 1947 and the sizes of the villages interpreted
upon the
Landsat TM
imagery,
proved
to be
highly
significant. Consequently,
most of the population growth
seems to be absorbed in the larger towns
and not in the
rural villa;2,es.
t~!, ..~.",..."I. h.§!.,.,... I;~..LQ.::;.§,§..;'~"L" . ,9..t . . . .~,.§l,.l.~h.D..i..?'.9..1j~.9,n,.:...
Several field surveys showed a rapid increase of the salt
affected soils in the depression.
The soils have been
classified as
Aridisols,
Entisols and Vertisols.
Most
important physical
factors for developing soil salinity are
the basin like topography.
the severe aridity,
the high
evaporation rate.
the shallow ground water table.
Most
important human factors are the inadequate drainage system
and the social stress at some locations were the human
pressure is high and the amount of available water is
somewhat restricted. This clearly shows that only a holistic
approach may result in some kind of success. Landscape as a
VIl . . 50
whole, with its natural and human components
should be
considered as valuable resource which can only be managed in
a integrated and ecological way (M.AntroP 1983).
j
I
Accurate mapping of the different degrees of salinization
could be carried out in a very satisfactory manner using
Landsat MSS (6th July 1981) and TM (12 July 1985, imagery
and the most simple interpretation techniques (Th.Ghabour &
L.Daels,1986). Photographic images of different spectral
bands were combined
producing diazo color composites. The
scenes were optical enl
to a scale of 1:100,000 and
interpreted visually,
Additional information was collected
from the stereos
when necessary. The
resulting map is presented in
3 and shows clearly the
increasing salinization in the per
ieal zone of the
depression as well as some 'high risk' spots in the central
part between some settlements.
5. The use of a GIS
MAP2 is the PC version of the Map Analysis Package (MAP)
adapted by De Dorschkamp (The Netherlands)
tsee A.Van den
Berg et.al. ,1984,1985). It is a low cost GIS and runs on any
PC XI or AI having at least 640 kb of RAM memory. It is a
raster based GIS which has a large set of procedures for map
comparison and spatial analysis.
Although the package was
primary developed for educational purposes and has therefore
a limited capacity, it became widely used in landscape
ecological research,
in landscape assessment and land
evaluation. The main drawbacks are the
input and output
procedures which are not users friendly.
Data must be
entered in raster format and no facility is available for
vector to ras~er conversion. Output of the maps is only
possible on a matrixprinter and not on the screen, which
gives a serious limitation for a fast interactive analysis.
Because only text screens are used the package can be used
on any PC configuration.
The price and the flexibility
determined this choice for the development of GIS on scales
of the governorate,
district and communal level.
Thus,
information from image interpretations maps. field surveys
and statistics could be combined.
The procedures
for spatial
analysis
proved
to
be
particularly useful.
As an example,
distance zoning was
carried out from the drainage and irrigation channels, using
terrain elevation data in a Digital Terrain Model
(DTM)
based upon
the
topographical
map
lfig.6).
Spatial
association with the interpreted salinity map (fig.7) and
with field survey maps proved to be significant in most
cases.
This
analysis supported
the
hypothesis
that
salinization increases with the distance from the water
channels and is more severe for drainage channels than for
irrigation channels.
6. Perspectives
Although d ital
imagery contains more precise information.
it is still very difficult to use this to
practical
1
work because 0f the restricted availability of the hardware
in developing countries. This is particularly true for many
small scale projects. Fortunately.
low cost software for
ciigi tal image processing becomes available for P(~ t()O, as
the PCIPS
tMyers~
1985).
Most procedures
of
image
enhancement can hardly be used for more ecological ways of
spatial analysis. Nevertheless, this kind of software may be
interesting for a more easy input of information in a raster
GIS. Also.
output of the GIS can be improved using digital
image software
to produce
coloured maps
of
better
readabili
than matrixprinted maps.
7. Conclusions
In the field
land evaluation and landscape assessment any
type of remote sensing imagery is an important and even
fundamental source of information.
Many enhancement and
interpretation techniques
(analog as well as digital) treat
that information with a very loose link with the complex and
holistic
reality on the terrain. The main reason seems to
t'e t.hat t.he necessary terrai.n i.nformation is not always
available and if so, it has seldom the expected perfection.
Meaningful interpretation with really practical use for the
local and regional planning authorities,
can seldom be
achieved without
an
integrated
landscape
ecological
approach. Techniques for spatial analysis available in GISpackages are even more useful then enhancement procedures.
Although an integration of
both is required. Anyhow,
the skilled human interpreter
remains indispensable.
8. Literature
Ant.rop M. \. 1982,\ - The "natural way" of visual image
interpretation for land classification and landscape
plannin~.
Toulollse,GOTA.Proc.lnt.Symp.ISPRS,Com.VII/l.
Archives vol.24-Vll/l,897-907
Antrop M. (1983) - Inventoring and monitoring of landscape
as a natural and cultural resource.
Paris.ESA,Proc.EARSel/ESA Symp. on Remote Sensing
Applications for Environmental Studies. 26-28 April 1982,
Brussels,105-113
Antrop M. {1986> - Structural information of the landscape
as ground truth for the interpretation of satellite imagery.
In: Damen. Sicco-Smit & Verstappen led.) : Remote Sensing
for Resources Development and Environmental Management.
Enschede,ITC,Proc. 7th Int. Symp.ISPRS.Com.VII,3-8
Ghabour Th. & Daels L. (1986) - Visual interpretation of
Landsat MSS images for the detection of salt-affected soils
<EI-Favourn. Egypt)
Myers H. & Bernstein R. (1985) - Image Processing on the IBM
Personal Computer. Proc.IEEE.vol.73,n~6,1064-1070.
Van den
A .. Lith J. & Roos J. (1984) - MAP: A set of
computer programs that
for input, ou
t and
transformation of cartogr
c data. Roskilde
Universitv.Proc. Ith Int.Seminar on Methodology in Landscape
Ecol
cal Research and Planning, theme I11,101-113
Van den Berg A., Lentjes P .. Lith J. & Roas J. (19854) MAP2 User Manual. Wageningen, Research Institute for
F()rl-.::strv and Larv:iscaJ:'e Planning HDe Dorsch~:amp".
PROCESSED AT FUCINO ON 17/04/86
KM 15
NASA
Fig.1.
The
Ju
ID:50497080~2
S/W ID. :011
TM5 3 177 040 2
sion of El Fayoum on Landsat TM band 3 (12
1985) ,
Lt)
co
~
~--
......
Lt)
....ra
{J
c:
ra
......
~
~
~
ffi
~
0
(!)
~
:;j
c:
0
1:J
Cat
VI
ra
..Q
VI
~
:1
ra
....c:
Cat
~
~ ......
....~
....
~~
Fig.2
Administrative divisions of El Fayoum Governorate
and the settlement shapes and sizes interpreted upon
Landsat TM.
VII ... 54
Fig.3
Soil salinity in El Fayoum based upon visual
interpretation of a Landsat TM colour composite.
1... 55
.4
. .5
Aerial
of the town El
Landsat I'M image (band
198.5 "
in 1956.
the t()wn E1
in
·.+
.+ •
00000000000000000000000000000000000000000000000000000000000000000000000000000000 +++
00000000011111 1111 U22222222233333333334~~444 .. " .. 455555SS5556666666666771777171711 . . +
••• 123456711901234567890123456711901234567119012345678901234567690123456769012345671190 +++
•• + 00000000000000000000000000000000000000000000000000000000000000000000000000000000
••• 000000000111111111122222222223333333333"" .... "4" .. "455555555556666666666777777777711 H+
. . . 123,,567690123,,56789012345678901234567890123.5671190123456789012345678901234567890 . . +
Ol)!
001
002 .
003
004
005
• H
002
003
00..
005
006
8
III.
III 1111
111160
. . . . . . 861110111111 •••• 1111111111111111
III1Uae611
. . . . . . 00881111111111111l1li •••• II II III iii l1li11.
111111&111118811&8
. . . . . . . . 11811&11.1111.11111111111111 •• 111111111111111 IIIl1l1BBBIIOeee
............ '!" ....... SIilIIllJilIilJia .. 1111 ...•. IIIIIBfilli1l1leeeOe
007
ooe
ClCI')
010
ou
012
013
01..
015
CU.6
011
018
OJ. 9
020
(121
oa:l
023
oa..
025
026
021
tl28
O~9
030
031
032
033
034
OJ5
036
037
Ollll
039
0 .. 0
0101
0402
04'
0 ....
. . . . . . . . . . . . . . . . . . . . . • ••• 111118 •• 11.111/ .IIIUlfJliIlIlIBleeeeOe ..... .
•••••••••• ;. ••••••• "1I1111111 •• IIIIII/IIII.111111188111180IHloeoo . . . . . . . . . . ,1111111
III1I1UIIIIIOeee&eeHU 11111111111I
......................... 4UIUIIIIII. illlllllllllleeell86ee4UIIIIIIIIIII
••••••••• --------- •• oUf ""IIII&IIS. SIIIUIIIIseeeeee8HeeetUIIIIII
.......... --------- •• ·t .. ueIlIlIIlBIIII.III1II1DilIIBlleeeeeeeeee,11111111
• ••• ••• ••• -------. . . . . . . . . . UliBIlIIIIIIII.IIII1I1II1Blllloeeseeeelll/lI f I l I I
• .........------. . . . . . . . . . . . . 01l1l11l1l1l.1I11111111111180e8&Oelll i 1IIIIIIIeoe
.........------••••••••••••• 01111111111. 811111111SIIeeeeeeelllllillee/ee
•••• ••• ... ------•••••••••••••• 188&111.11 •. ••••• ""e8., •• 11I1111118088
••• -----.-----••••••••••••••••• 011.1111111. III .• 1111111.111/111111111.1111
------------- •••••••••••••••••• 111111.1 •••• _I.IIII •• SIIBIIIIIIIII
-----------••••••• • •••••••••• "0 •••••••••• 11 ..• 8.IIIIIIIIIIIIIIUI
-----111-----•••••••••••••••••• 00 •• 1/11' ••• 1.1. ,,1.811111Ia.1
----11111----•• • ••• ••• ..... ••• .... ••• •• //111 •••• 1.111 . ....... ... 1
.. _---, / 1111-- •••••••••••••••••• •• .... , /1///8.11111.1.111118111.1
xxxxxx---- ... -- .. IIIIII-- . . . . . . . . . . . . . . . . . . . & 1IIIIIIIIIe08eOI
JlI.1II
•••••••••••••••••••••••• U1I1I881101l1l.
xxxxxxx-------IIIIIIIII ••••••••••••••••• •• ·,1111111111111
xxxxxxx:-------III/IIIIII.· •••• • ...... ~ ••• •• ··111111111111
XXXXXXXQ--- ... /IIIIIIIIIII . . . . . . . . . . . . . . . . . •••• •• ·'11111111
xxxxxXXXX----IIIIIIIIIIII . . . . . . . . . . . . . . . . . . •• .. • • •• /1111111
xxxxxxxxXXX---IIIIIIIIIIIIIIII ••••••• • ••••• ••• •••• /1111II
lIXXXXXXXXXXXXXXfllll1 1111111111' ,:. •••••••••••••••••• 11111
xxxxxxxXXXXXXXIIIIIIIIIIIIIIIIIIII . . . . . . . . . . . . . . . . . . . . . 111
XXlCXXXXXXXXXXXXXIIIIIIIIIIIIIIIIIIIIIIIIIIII···.·····.111
XXllXXXXXXXXXXXXXl1 /1111111111111111111111 111111 ••••••• 1111
XXXXXXXXXXXXXXXIIIIIIIIIIIIIIIIIIIIIIII / 1I11111111111I11111
lIXXXXXXXXXXXXXlllllllllllllllllllllllllllll1 f I1I1111111II111
IIXXXXXXXXXXXlllllllllf
NIIII
",IIXXXXXXXXXIIIIIIIU
I I I IIIIIIXXXXX'XXII II II II
lillI/lillI/lXXII/I II II " i
III 11111111111111111111111
11111111111111111111111111
1111111111111111111111111
111111111111111111111111
111111111111111111
0 .. 5
I.
001
002
003
004
005
006
007
008
009
010
OU
012
013
014
015
016
017
OUI
019
020
021
022
023
006
007
eee.........
........ KKKKKK ....................... ee.eeeeee •••••••••••••••••••••••••••••
••• •• ....··XKJ( ... KKKKXX ............... eeeee ••• ee ........................ .
... • ••••• XKKKXKXXXXX)(H ............ eeeeeeeeee ..........................
...........XJUOOO()(X)()(KK ......... eeeeeeeeeee •••••••••••••••••••••••
••••••••• X}UUO()(X)fXx .......... ee&&eeeese .............................
•••••••••XKKKKX)(XKK .......xxxeeeeeeeoeee ••••••••••••••••••••••••••
011
012
013
o1 ~
··.XXXKXXXIIKXXKXIU(IIXXXKIIXK08ee8e~eeeeKXIIXI( •••••••••••••••••••••••
015
•• XKXXXXKKKKXKXXXXX.· •••••• e.eeaeeeeKxxKx •••••••••••••••••••••••
•••• .KHKKKXKHKXXHKXXx: ........... &8eexKee •• e ..........................
•••• xKKnx.)OOuCX}OCxx ...........HJO()(XKXXX)u<e .... e ••••••••••••••••••••
KHKJUCXHKKKKX •••••••••• XKJ(XXHKHKKx ••••• e ••••••••••••••••••
016
017
.1II...
018
019
...... X.XXXXKK ............... KKKXXKKXKKKKxKx •• ee . . . . . . . . . . . . . . . .
020
............................. XKXX)(KXXXXK)()(KXKHKK.e ............. ..
021
022
........................ III . . . . . . XK.KXXKXXKXXXK)(HX ••••••• eoee0:0981.
•••••••••••••••••••••••••••• X)(KXXXJUOU()(KKxee ••••••••••••••• e ..
023
021e
025
026
027
028
029
030
025
026
021
028
029
•••••••
OJ3
. 03,.
035
036
037
03111
039
040
0101
042
043
III· •• •• ••• •• •• ••• ••• • •••• •• .... XKKKKKKX)OUCKK)(KeOe8..
•• •••• • ••••••••• • •••• • ••••• • •••••••••• KXXJCKKXHKKXXXKK8KKK
..e.. e
. . . . . . . . . . . . . . . . . . . . . . . . . . . . .·......... XKXXXXXJUUOOUOU(X)()(HK
............................... XXXXXXXXKKK)(HHX):{XXx)(HKKHKXH
••••••• • •••• •• ......... ••• .... ·xxxxxXXXXXXHXKXXXXKXHKKKKXXH:
............................ XXXXXXXXXXXXXXXXXXXXXKXXXXXIIXX
•••••••••••••••••••••• IlI • • • • • • XXXXXXXXXXXXXX10UtXXKJOfx:KXXKK
030
031
03l
. .......... .
.ID ...........
••••••• IIII • • • • XH ........... ee .... eeee& .............. ..
........ • ••• xKx .. e..................... eeeeee.......... • ••••
ooa
009
01 0
0~4
031
032
033
034
035
036
037
03111
039
040
041
042
043
044
045
-e
••••••• 8886 •••••••••
.....................
.......................... 88........
..................................xxxxxXXXXXXKKKKKHKXXXKlClOCK
.1III.1II'11111 .................... BI . . . . . xxxxxxXXXXxxHlounoooouonOHe
1IIII • • • IIIII ............................... xxxxxxxxXXXxxxlounuunUUt"'K
••• IIIIIIII •••••••• IIII • • • X •••• a • • • • X .. X••••••• XXXXXX10UUUOUUOUC
IIIIIIIIIIIII ••••••••• XXX ............................. XXXXXHX HKKXXM
11111 fllIlIlIlIl···IIXXX
/I IIIIIIIIIIIIIIIIIIXXX
III IIIIIIIIIIIIIIIIIIXXX
III 1111111111111 1111111'11.11
11111111111111111111111111
11111111111111111111111111
111/111111111111111111111
111111111-11111111111111
1111111 .........111
0440
0~5
001
002
003
004
005
006
001
·008
009
010
OU
012
013
014
015
016
017
018
019
020
021
022
023
024
025
026
027
028
029
030
031
032
033
034
035
036
037
0:4111
039
0 .. 0
041
042
043
044
045
. . . 00000000000000000000000000000000000000000000000000000000000000000000000000000000 +++
..... 00000000011111111 11 22222222223333333333444444444455555555556666666666 71717171778 .. H
.... &2345671119(1234567891)1234567890123 .. 56111190123456789012345678901234567111901234567890
.... 00000000000000000000000000000000000000000000000000000000000000000000000000000000 ....
+ ... 00000000011U111111222222~222333333333l .. 4"4 .. " .. 444555S5S5555666666666677771717178 +++
. . + 123456711190123456?890123456789012345678901234567e901~34567e9012345678901~34567890 . . .
liP HAHIT. DIIA1SPD
IIUDER I D1 .. tanc.. aonln. d .... nhill • aain drain chann.. l IU Hinya
$CALlE
100
I
COVERAGE COVERAGE
311'HIiIOi. VAWI LAIiIL
.11
CILLS
-Ii
AP HAHI. XSALDRAJ
HUllER , CoabinaUon d1 .. tance lI!ontnJII or draina.:" with "alinity in III H1n111
100
SCALI
COVERAGE COVERAGE
SYMBOL VALUI LABEL
CELLS
+B
-B
.:0"
B
0
5
lIXXX
1111
1111
10
20
30
100
50
60
70
80
90
100
.
D;~i~~~~ ~h~~~i
500a
ll<a
2 It"
3 It..
'" Ita
5 k ..
6 ka
1 It.
8 kill
9 ka
10 'UI
....
1506
106
373
144
86
65
22111
2ll
86
51
168
47
581
41.1113
1.28
10.36
4.00
2.39
1.1111
6.33
5.92
2.39
1.42
4.67
l.:U
16.31
.00
2.20
11.1111
6.88
4.11
3.10
10.89
10.17
4.11
2.4"
8.02
2.2.
2111.03
lIeee
XXKI(
B
0
2
3
XXXX
XXX)(
6
7
1111
~
. ....................
bulld .. r ..a
non lIIal • 1 Itll
aod "el • 1. 51t"
non alll 1-2. 5k ...
"od DIll ~-~.5k ..
non " .. 1 , 3 k ..
mod .... 1 • 3 Ie ..
very .... 1 ) 3 Ie"
1511
13;<
602
437
489
92
99
15
~23
41.97
3.67
16.72
12.14
13.58
2.56
2.75
.42
6.19
.00
6.32
~11.82
~0.92
23.41
4.40
4.74
.72
10.67
Fig.7
.6
Part of the MAP2 GIS of El Fayoum : commune El
nya. Distance zoning around the main drainage
channels, related to the topography.
Part of the MAP2 GIS of El Fayoum
commune El
Minya. Combination of the interpreted salinity map
(Landsat TM, 1985) and the distance zoning around
the main drainage channels.