CC BY
17GEOSTATISTICS, REMOTE SENSING AND GIS
Ibrahim Yahiaoui, Abd-el-Kader Douaoui, Ahmed Ziane
DIGITAL MAPPING OF LANDSCAPES BASED ON SOIL MORPHOLOGY IN THE PLAIN OF LOWER-CHELIFF (ALGERIA): APPLICATION OF REMOTE-SENSING
University of Khemis-Miliana, Laboratory of Agricultural Production & Sustainable Valorization of Natural Resources, Khemis-Miliana, Algeria, ibyahiaoui@,gmail.com ,
Tel: +21327423632
Abstract. The Plain of Lower-Cheliff is one of the most studied north Algerian areas in terms of soil and water dynamics either in space or in time but it is very poor in term of landscapes mainly those related to vegetation. Here in this brief study we used a digital elevation model to detect the soil morphology of our study area mainly slopes and altitudes then to extract and classify the different landscape patterns using the method of supervised classification applied on a satellite imagery of LANDSAT TM type to present the spatial distribution of the main morpho-landscpaes in a digital spatial map. The resulted classes from the image are mainly: cereal -crops, tree crops, garden-crops to the east side of the area also, halophytes and bare soils to the west side. The digital elevation model shows that the variation of elevation is not that important as the slope but it is still the one that controls the distribution of landscapes in the plain according to the overlay, elevation/landscapes: halophytes and bare soils locate in low altitudes commonly called depressions following the soil salinity concentration at this level, other crops locate in mid to high altitudes away from soil salinity that endangers agricultural practices in the area, finally we can say that the spatial distribution of any landscape in the area is directly related to soil dynamics and characteristics.
Keywords: Landscape, soil morphology, salinity, satellite image, digital map, Lower-Cheliff.
INTRODUCTION
The study of landscapes is regarded as an important information, for the evaluation of surface layers dynamics and the comprehension of the independent factors, which are responsible for any kind of degradation that can occur. The knowledge of these layers and their main component on a large scale remains always difficult and often not providing sufficient details to know their properties and to understand the operating mechanisms of the soil such as salinisation (Mulder et al,2011).
The plain of Lower-Cheliff located in the north west of Algeria is known with a variable dynamics of its surface layers under the direct influence of soil salinisation which occurs on different levels (Khan et al, 2001) (soil, vegetation) either throw irrigation (secondary salinisation) or coming from underground layers (primary sali-nisation).
The realized works in the area has all been unanimous to say that the soil salinisation which does not cease extending spatially and increase temporally, according to (Douaoui et al, 2006) is the leading cause of this degradation which appears by the deterioration of the vegeta-
tion cover and the water and soil resources (Fernandez et al, 2006).
McBratney et al, (2003) proposed a framework for the predictive digital mapping of the ground surface by considering the environmental variables (morphology, vegetation, lithology...). In the same direction, Boettigger et al, (2008) showed efficiently the utility of the satellite images to determine the characteristics of the surface layers named "landscapes" in arid regions with the presence of a moderate vegetation cover.
This work, aims to characterize and to map surface layers mainly landscape patterns; to know their spatial distribution in the area and their relationship to the digital data and the extracted parameters from the used digital elevation model (DEM).
MATERIAL AND METHODS
Material studied
We used two types of digital data downloaded from (http://earthexplorer.usgs.gov), the first is a satellite image of LANDSAT TM5 type with seven spectral bands covering the whole study area (Table 1) the second type is a digital elevation model which expresses
the two morphologic data (elevation) for the stud- age was taken during a vegetation season ied region (figure 2). We validate the chosen pat- where the red color refers to the vegetal cover terns in the study throw a field mission. The im- in the area (spring) 04/08/2013
Figure 1 - The LANDSAT TM 5 image of the study area
While most of the satellite data are provided with digital accounts geometrically corrected and to better map each targeted pattern in the area we proceed to an atmospheric correction in order
The used DEM was established in 08/11/2013 with 24.5 meter of resolution and was converted into elevation values in meter. Study Area
The Lower-Cheliff known as "The Saline ground" (Douaoui et al, 2006) is located in the
to prevent any atmospheric noise (Browing & Duniway, 2011) and to convert the digital accounts into reflectance values.
north west of Algeria at 220 km from the capital Algiers (figure 3), it is the west part of the grand Cheliff basin and extends on more than 50000 hectare.
Table 1 - Characteristics of the used satellite image
Scene ID Band Combination Acquisition Date Resolution
LANDSAT TM5 Red -^B4 (NIR) Green -► B3 (R) Blue -► B2 (G) 04/08/2013 28.5 meters
NIR : Near Infrared, R: Red, G : Green
Figure 2 - The extracted DEM for the study area
Figure 3 - The location and delimitation of the Study Area
Our area is characterized by a semi arid climate with hot and dry summers, cold and warm winters also by various types of soils which are in most clayey to the east of the plain with low percentage of loamy and sandy to loamy to the west side, cropping systems are focused in areas where salinisation is low or even null such as Ouarizane, Djediouia and some parts of H'madena, there are places where plants life is
critically impossible such as the extremly saline depression of Gaa where we can find a stretched presence of halophytes, other zones know instable agricultural practices threaten also by the harsh weather there, the area lies on a large extension of degradable soils named bare soils under the continuous influence of salinity concentration in space and time (Mokhtari et al, 2012).
METHODS A tmospheric correction This correction is based on converting digital number of each pixel of the image into spec-
IA =
Lmax - Lmin
tral radiance (USGS, 2001) then to convert the radiance values to surface reflectance in order to eliminate the atmospheric noise on the surface patterns:
-:-;—— x (Qcal - Qcalmin) +
.Qcalmax - Qcalmin/
Lmin............(1)
where the LMINs and LMAXs are the spec- 255, respectively, and QCAL tral radiances for bands at digital numbers 1 and ber (Chander et al. 2009).
digital num-
7T.L^.d'
ESUNÁ.cos(SZ)'
(2)
where
p = planetary TOA reflectance [unitless] (Chander et al. 2009)
spectral radiance at the sensor's aperture in [W m -2 sr -1mm -1]
d2= the square of the Earth-Sun distance in astronomical units = [1 - 0.01674 cos(0.9856 (JD-4))] 2, where JD is the Julian Day (day number of the year) of the image acquisition
ESUN^= mean solar exo-atmospheric irradiance in [W m -2mm -1 ]
SZ = sun zenith angle in radians when the scene was recorded.
We processed the data throw ENVI 4.8 platform using specific tool for LANDSAT calibration.
Classification Process We chose to use the supervised classification because of our previous knowledge of the field and because we are targeting landscape patterns we adopted the algorithm of "Maximum Likelihood" as recommended by Girard & Girard, 1999, while it deals with pixels that belong to the same pattern in a narrow error possibility, the validation was done by calculating the KAPPA coefficient which was more than 0.99 (99.17 % precision). This allowed us to have a sum of classes that cover the main surface layers in the area (Fig.5) DEMprocessing
According to Arrighi et Soille, 1999, we need to extract the elevation from the used DEM
using the "Spatial Analyst" tool in ArcGis software where we will have an independent map for elevation with values in meters and clipped as for the satellite image. The Overlay image/DEM To answer the problematic we needed to combine the results of classification scheme with the extracted morphologic parameters, this was done using ArcGis platform giving the viewer the possibility to see in 3D the spatial distribution of the morpho-landscapes in the area.
RESULTS AND DISCUSSION Elevation extraction The aspect plain was given to the area because of the low variation of morphology especially in term of elevation (figure 4) The plain of Lower-Cheliff has low altitudes with a predominance in the depression of Gaa (30-40 m) and some surfaces near to Hmadna (45-55 m), Djediouia and even Ouarizane (50-60 m). In the east of the plain and on its peripheries, elevation increases to reach the 140 m on the level of the hill of division. The Hill of Benziane reaches altitudes which vary between 70-120 m. In a general, Except the zone of the hill of Benziane, elevation is low in the western part of the plain and increases proportionally towards the east.
Figure 4 - Elevations map of the Lower Cheliff plain
Classified Map
Nine classes were obtained, where four classes represent the vegetation (Halophytes, Trees, cereals, garden), two classes for the agricultural vegetation (dense and average dense), two classes for the soil (Bare and plowed soils),
two classes for water ( the free water surface and the Cheliff wadi), and a class for the sebkha.
These classes are obtained during a spring period where all patterns of landscape can be detected especially vegetation.
Figure 5 - Classification scheme of the main landscapes in the study area
Overlay DEM/Image
To see the natural distribution of the area based on terrain morphology we overlaid the dig-
ital elevation model on the satellite image (figure 6.1) giving a 3D visualization of the study ground there.
Depression of GAA
Figure 6.1 - 3D visualization of the study area morphology
place in low and mid altitude where salinity is very low or null, the moderate altitudes in the area include an heterogeneous distribution between Trees, cereals, plowed soils and even a small part for bare soils (figure 6.2). The Free water surface in the plain called "Merdja" is the result of water collection due to soil structure which is clayey and compact, it is located in mid altitudes.
Overlay Map/elevation Most of the landscapes are on similar elevation levels such as bare soils and halophytes plants which occupy in general the very low altitudes called "depression" following the high levels of salinity in the area, in second position plowed soils and garden crops are mostly in mid altitudes where farmers find good conditions to practice there, the famous agricultural cropping systems such as Tree cropping and cereals take
Figure 6.2 - Digital map of landscapes morphology in the area (3D Visualization)
CONCLUSION area throw classification process then to see
The combined use of a satellite image and their distribution at different levels of eleva-DEM was of big benefit by giving us the possibil- tion in the area, where we can define a limitary to locate the main landscapes patterns in the tion of their extension in term of altitude Z (table 2)
Table 2 - Limitation in altitude for landscape classes
Landscape Class Altitude limitation (meter)
Plowed Soils 65 > Z > 40
Bare Soils 40 > Z > 18
Cereals 78 > Z > 35
Gardens 70 > Z > 55
Halophytes 55 > Z > 18
Tree 122 > Z > 65
This results can be extended by confronting phology and digital processing by creating a the digital data either from the image or the DEM database that can help in studying the vari-to field sampled data under the purpose of estab- ability of the surface patterns in space and lishing a relationship between, the field, mor- time and to see which influence has the topography on them.
REFERENCES
Arrighi., A et Soille., P., 1999. From scanned topographic maps to digital elevation models. In Proceedings of Geovision, p196.
Boettinger, J.L., Hartemink, A. McBratney, et Mendonca., S, 2008. Landsat Spectral Data for Digital Soil Mapping, in Digital Soil Mapping with Limited Data, A.E., Editors. Springer. p. 193202.
Browing., M et Duniway., M., 2011. Digital soil mapping in the absence of field training data: A case study using terrain attributes and semi-automated soil signature derivation to distinguish ecological potential.
Chander GD, Markham BL, Helder DL (2009) Summary of current radiometric calibration coef fi cients for Landsat MSS, TM, ETM+, and EO-1ALI sensors. Rem Sens Environ 113:893
Douaoui A., Nicolas H., Walter Ch., 2006. Detecting salinity hazards within a semiarid context by means of combining soil and remote-sensing data. GE0DERMA,134(2006).
Fernandez-Buces N., Siebe C,. Cramb. S, Palacio J. 2006. Mapping soil salinity using a combined spectral response index for bare soil and vegetation: A case study in the former lake Texcoco, Mexico. Journal ofArid Environments 65 (2006) 644- 667.
Girard M.C. & Girard C.M., 1999. Data processing ofremote sensing. Dunod, Paris, 529 p. + CD-ROM.
Khan N. M., Rastoskuev V. V., Shalina E., Sato Y., 2001. Mapping salt-affected soil using remote sensing indicators. A simple approach with the use of Gis Idrissi. 22nd Asian Conference on Remote Sensing, 5-9 November, Singapore, 2001.
McBratney., Hartemin., A., Mendonca- Santos., M., 2003. Mapping with Limited Data, Editors Springer. p. 193-202.
Mokhtari D., Douaoui A., Yahiaoui I., 2012. Geomatics use in the evaluation of surface qualities degradation in saline area (The case of the lower Cheliff plain). Journal of Energy Procedia. Issue 18 (2012). 1557-1572.
Mulder V.L., de Bruin S., Shaepman M.E., Mayr T.R., 2011. The use of remote sensing in soil & terrain mapping - A review, Geoderma Journal/Elsevier, Issue 162 (2011) 1-19p.
USGS., 2001. Landsat 7 science data user's handbook. USGS, Reston Zobeiry M, Maid AR (2000) An introduction to remote sensing technology and application in natural resources. University of Tehran, Tehran.
РЕЗЮМЕ И. Яхиауи, А.К. Дуауи, А. Зиане
СОЗДАНИЕ ЦИФРОВОЙ КАРТЫ ЛАНДШАФТОВ НА ОСНОВЕ МОРФОЛОГИИ ПОЧВ РАВНИННОЙ ЧАСТИ НИЖНЕГО ЧЕЛИФА (АЛЖИР) НА ОСНОВЕ ПРИМЕНЕНИЯ ДИСТАНЦИОННОГО ЗОНДИРОВАНИЯ
Университет Хемис-Милиана, Лаборатория сельскохозяйственного производства и устойчивого валоризации природных ресурсов, Хемис-Милиана, Алжир, [email protected] , Тел : +21327423632
Равнина Нижний Челиф является одной из наиболее изученных Северо-Алжирских районов с точки зрения почвенной и водной динамики в пространственном или временном аспектах, но недостаточно изучена в плане ландшафтов, главным образом связанных с растительностью. Здесь, в этом коротком исследовании, мы использовали цифровую модель рельефа для определения морфологии почвы на исследуемом участке, в основном, склоны и высоты, с тем, чтобы затем получить и классифицировать различные типы ландшафтов с использованием метода контролируемой классификации, применяемой на спутниковых изображений типа LANDSAT ТМ, чтобы представить пространственное распределение основных морфо-ландшафтов на цифровой пространственной карте. На основе анализа изображения выделены следующие классы: зерновые -с/х культуры, древесные культуры, садовые культуры в восточной части, солянки и голые почвы в западной части. Цифровая модель рельефа показывает, что изменение высоты не так важно, как уклоны, но оно по-прежнему контролирует распределение ландшафтов на равнине в соответствии с наложением сочетания высот / ландшафтов: солянки и голые почвы находятся в низких высотах, в так называемых депрессиях, следуя за степенью засоления почвы на этом уровне, другие культуры находятся в средней части на больших высотах, вдали от засоленных почв, которые ставят под угрозу сельскохозяйственные работы в этой области, и, наконец, мы можем сказать, что пространственное распределение любого ландшафта в районе, непосредственно связано с динамикой почвы и ее характеристиками.
Ключевые слова: ландшафт, морфология почвы, засоленность, спутниковое изображение, цифровая карта, Нижний Челиф.
ТYЙIН
И. Яхиауи, А.Э.К. Дуауи, А. Зиане
Т0МЕНГ1 ЧЕЛИФТЕГ1 (АЛЖИР) ЖАЗЬЩ ЖЕРЛЕРДЩ ТОПЫРАГЫНЬЩ МОРФОЛО-
ГИЯСЫ НЕГ1З1НДЕ ЛАНДШАФТАРДЬЩ САНДЬЩ ЖYЙЕДЕП КАРТАЛАРЫН К¥РАСТЫРУ: КАШЫЩТЫЩТАН БАЙКАУДЫ ПАЙДАЛАНУ
Ауылшаруашылъщ вндiрiс жэне табиги ресурстарды турацты валоризациялау зертханасы, Хемис-Милиана, Алжир, [email protected] , Teл : +21327423632
Теменп Челиф жазыктыгы кещстшп немесе уакытша буындардагы топырак жэне су динами-касы т^ргысынан солтYCтiк Алжирдщ барынша зерттелген аудандарыныц бiрi, алайда еимдште-рмен байланысты ландшафтар т^ргысынан жеткМкп TYрде зерттелмеген. М^нда, кыска зерттеуде бiз бакыланатын учаскеде, непзшен жарлар мен к¥здардыц топырагыныц морфологиясын аныктау Yшiн сандык ЖYЙедегi Yлгi колдандык. вйткеш, б^л кешннен LANDSAT ТМ типп спутниктш бей-нелерде колданылатын бакыланбалы жiктеу эдiсiн пайдалану аркылы ландшафтыц TYрлi Yлгiлерiн алып, жiктеу, сандык ЖYЙедеге кеиiстiк картасында негiзгi морфо-ландшафтардыц кеиiстiктiк бвлiнуi туралы TYсiнiк алу Yшiн кажет. Бейнелер нэтижесiнде алынган санаттар негiзiнен мынадай: астык-ауыл шаруашылык дакылдар, агаш мэдениетi, учаскенiц шыгыс жагындагы бау-бакша мэде-ниетi, батыс бвлiгiндегi сорац жэне такыр жерлер.
Бедердщ сандык ЖYЙедегi Yлгiсi еиiс секiлдi биiктiктiц взгеруi соншалыкты мацызды емес екенш, бiрак ол YЙiндi, YCтiрт/ландшафтарFа сэйкес жазыктыкта ландшафтыц Yлестiрiлiмiн б^рын-гыша бакылайтынын бiлдiредi: сорац жэне такыр жерлер эдетте ойпат деп аталатын темен биiктiктерде сортац жерлерде орналасады, баска дакылдар ортада жоFары биiктiктерде осы облыс-тыц ауыл шаруашылык ж^мыстарына кауш келтiретiн сортац жерлерден жыракта орналасады. АудандаFы кез келген ландшафтыц кещспкл орналасуы топырак динамикасымен жэне сипатымен тiкелей байланысты.
Кiлттi свздер: ландшафт, топырак морфологиясы, сортацдану, спутниктiк бейне, сандык жYЙедегi карта, Теменгi Челиф.
