 |  |
| Bildiri Özetleri | |
| Ana Sayfaya Dönüş |
| Back To The Main List |
|
Çetin Arcak 1, Şenay Özden 1 Yusuf Kurucu 2, Ünal Altınbaş 2, Mustafa Bolca 2,Tevfik Türk 2 1 Soil & Fertilizer Research Institute, Ankara, Turkey 2 Ege Univeersity, Faculty of Agriculture, Department of Soil Science, Izmir, Turkey Abstract Among geographic regions apt to mass movement is Turkey that is liable to its diverse forms. The study aimed at generating a GIS and RS aided model that could be very useful tool in estimating mass movements leading to soil degradation and mapping susceptible zones. Within this context, expediently generated soil maps were given special emphasis for their usage in degradation studies related to mass movement and provision efforts. Direct mapping methodology and Heuristic technique were applied in the study. Since direct method requires in large part a geomorphological approach, following a photo-interpretation based semi-detailed soil survey, a geopedological map of the study area was generated. Resulting mass movement features which were tried to delineate on the map are fans originating from mud flows; stone stripes located in relatively shallow gullies; debris and talus forms accruing from the disintegration and the slump of the rock substratum. Model makes estimation of mass movement hazard zones possible through the manipulation of potential risk factors such as slope, geology, land cover and soil properties regardless of mass movement features and types. For the purpose of determining hazard severity rating and risk factors, stationary factors such as slope steepness, properties of the rock substratum, land cover formations and taxonomic soil groups were used. Aforementioned factors were subjected to a series of queries using Geo-media 3.0. Slope was considered dominant factor and relations between slope-geology and slope-taxonomic soil groups were found out. After a relational query, five groups were generated. These groups were intersected with land cover producing a new layer for a new query. All possible combinations of the four factors were tried and 26 appropriate combinations were obtained. These files were classified in descending order of susceptibility. Finally hazard severity classes were established as follows; very high, high, moderate, low, non or uncertain. Introduction Among natural hazards, mass movements are the most frequent and have the largest geographic distribution, landslides alone, for example, cause greater loss of human lives and economic goods than any other single natural hazard. Moreover, they may result in unrepairable damages in irrigation and drainage networks, telecommunication and highway systems. Some types of mass movement displace impressive volumes of soft or slightly consolidated rock materials over large, regional surface areas and long distance. Mass movement manifests itself, for example, as landslides; solifluction or mudflow, a downward movement of water-saturated soil, terracette and creep, a relatively slow downward movement of soil; slump, a downfall of disintegrated rock materials. The spatial distribution of the mass movement prone areas is strongly controlled by the rock-soil-slope-land use complex (Lopez and Zinck, 1991). Method The study area is located in Senirkent township of Isparta province encompassing an area of about 300 km 2. Direct mapping method and heuristic technique were used in the generation of hazard maps (Barredo at all, 2000). Since direct method requires to great extent a geomorphological approach, geopedological map of the study area was generated. To this end, an implementation program of two phases was adopted. First phase included reconnaissance of the area, interpretation of the aerial photos and satellite imagery with the emphasize on the geomorphology and morphodynamic characteristics of the area and developing field legend. Second phase involved the generation and the acquisition of thematic maps such as slope, lithomorphology, land use and geopedology, determination of sample areas, soil survey activities including profile descriptions and mechanical, physical and chemical analyses, definition of risk factors, queries and modeling (Figure 1.) Processing and Results 1/35 000 scale 26 panchromatic aerial photos of 1991 were interpreted by making use of satellite imagery, geology and topographic information. 13 profiles representing sample areas were dug and sampled one the basis of genetic horizonation. For the purpose of final delineation, entire boundaries designated on the photo-interpretation map were checked, rectified where required, validated and extrapolation was full filled. In order to generate attribute data, final soil map was digitized (Figure 2.) using Geo-media 3.0 software. Model inputs produced from the map for further processing are slope gradient by landform, taxonomic soil groups and lithomorphology. Digital Terrain Model (DTM) was derived from 1/25 000 scale topographic data through Microstation-Terrain Analyst Module. Land use charecteristics of the area were provided by a study of 4,5,3 bands combination of landsat 5 TM imagery taken in August 1998. Required rectification, image processing, enhancement and supervised classification procedures were executed using image Analyst Module.  Geopedological approach: "land form" was used as mapping unit in the geomorphology-based soil map, and taxonomic soil compositions of the each landform were described. In the methodology being applied, study area was differentiated by landscape properties (Mountain, plain, piedmont plateau), then each landscape was classified in itself by relief, lithology and landform being the smallest homogenous units. Observed morphodynamic features were tried to delineate within the landform in which they were described. Resulting morphodynamic features identified are the fans originated from mudflow, slumps bringing about debris-talus forms and stone-stripes formed by the conglomerated stones and rock fragments in the relatively shallow gullies (Figure 3).  The X-ray diffraction analyses revealed that the composition of clay minerals of the selected horizons in some profiles as follows;35,87 % illite, 35.23 % kaolinite, 18,21 % montmorillonite, 6.92 % vermiculite or chlorite. These minerological values can be lined in order of amplitude as illite > kaolinite > montmorillonite > vermiculite > chlorite and correlate with liquid and plastic limits. Modelling:For the purpose of defining risk factors and hazard severity rating being the basis for mass movement hazard zoning, stationary factors were classified in ascending order of degradation susceptibility as follows:  For soil susceptibility rating; rational weighed soil group was selected from taxonomic composition of each landform. In addition, pedologic evolution and the properties of taxonomic subgroups were taken into account for this purpose. As to lithological rating; resistance of the geological material to degradation was ranged in descending order as above. In this order, lime stones of different origin (pelagic, neritic and red limestone) were incorporated into same group, minor lithological units in the formation were included in this major formation. For the susceptibility of slope gradient; slope ranges of previously conducted researches of similar nature were considered. For land use rating; dominating land uses and land covers identified through the image processing of Landsat TM 5 satellite imagery were classified as regards their soil protection capability as well as other regional factors contributing to this capacity. Attribute data of soil, geology and slope from above mentioned four factors were gathered under one file which enabled the generation of one layer of three factors. In this layer, several queries were executed by considering the susceptibility to mass movements. For the purpose of determining hazard severity rating and risk factors, stationary factors such as slope steepness, properties of the rock substratum, land cover formations and taxonomic soil groups were used. Aforementioned factors were subjected to a series of queries in Geo-media 3.0. Slope was considered dominant factor and relations between slope-geology and slope-taxonomic soil groups were found out. After a relational query, five groups were generated. These groups were intersected with land cover producing a new layer for a new query. All possible combinations of the four factors were tried and 26 appropriate combinations were obtained. These files were classified in descending order of susceptibility. Finally hazard severity classes were established as follows; very high, high, moderate, low, non or uncertain (Fig. 4)  Bare rocks, overgrazed and weak pastures on the Lithic Xerorthents and Lithosolic units with slope gradient of 50 % formed on the underlying lime stones and flyschs are expressed as very high susceptible zones. Due to its increased lithological resistance in proportion to limestone and flysch , dolomitic units having pastures, bushes and shrubs land cover are classified as highly susceptible zones, one category below the very high. Level or almost level (0-8 %) Typic Xerofluvents and Vertic Xerofluvents of alluvial origin under irrigated and non-irrigated agriculture, bushes and shrubs, and some colluvial soils are designated as non or uncertain. Conclusion It can be inferred from the mass movement hazard zoning map that due to its proximity to the two very high susceptible gullies, Senirkent is the most vulnerable settlement in the study area, and any prospective mass movement threat most likely originate from the very high susceptible zones on the upper summits through the gullies and its tributes. From the stand point of analytic data, there is significant amount of montmorillonitic clay accumulation in the profile as from the depth of 80 cm in the soils within the close vicinity of Senirkent which is classified as moderate. Soils of this surroundings have also high plastic and liquid limits. In the case of heavy or extreme rainfall, depending on the water load and slope gradient, if soil moisture content reaches PL, slides will likely occur. Similarly, when soil is saturated with water up to its LL value, fluction might take place (Tarhan,1991). With the methodology followed in this study, the importance of geopedological approach in mass movement-related studies and the generation of a model adaptable to other regions were intended to emphasize. The study is not independent of each other in terms of content, however, it may be approached from two perspectives, one is the production of time-consuming geomorphological soil map, the other perspective is the cartographic model generation which could have been produced from data other than that introduced here. In this context, it is possible to express that both a) have distinct manner b) may calibrate each other c) can be considered complementary of each other. Model introduces mass movement hazard zones by taking factors of potential risks such as slope, geology, land cover and soil properties into consideration regardless of mass movement features and types. The fact that geomorphological approach enables the mapping of landforms characterized by mass movements provides a good understanding of the mass movement types being formed and the likelihood occurrences. If contributed by analyses, it makes interpretation and estimation towards prospective or past mass movements easier. References . Barredo, J., Banvides, A., Hervas, J. and Van Westen, C.J.. 2000. Comparing heuristic landslide hazard assesment techniques using GIS in the Tirajana basin, Grana Canaria Island, Spain. International Journal of applied earth observation and geoinformation . ITC. Enschede-The Netherlands. . Lopez, H.J, Zinck, J.A., 1991.,. GIS - assisted modelling of soil-induced mass movement hazards: a case study of the upper Coello river basin, Tolimo, Colombia. International Journal of applied earth observation and geoinformation . ITC. Enschede-The Netherlands. . 3. Tarhan, F. 1991. Doğu Karadeniz Bölgesi Heyelanlarına Genel Bir Bakış. Türkiye 1. Ulusal Heyelan Sempozyumu. S-38. Karadeniz Teknik Üniversitesi , Trabzon. Legend of Figure 2
 |
