|
|
| Bildiri Özetleri | |
| Ana Sayfaya Dönüş |
| ISD Ana Sayfası |
|
REMOTE SENSING AND GEOGRAPHIC INFORMATION SYSTEMS IN THE VICINITY OF ANKARA ÇUBUK DAM LAKE Şenay Özden Soil and Fertilizer Research Institute Nuri Munsuz Ankara University, Agricultural Faculty, Soil Science Department ABSTRACT Due to stimulating topographic structure, erosion is a crucial problem in Turkey. In this study, objective was to determine, through quantitative modelling using RS and GIS techniques, changes resulted from occurance of erosion by years. Materials used in the study include aerial photos of 1969 and 1995 at scales of 1:21000 and 1:4000 respectively, topographic, soil and geological m aps of 1:25000 scale, Landsat 7 images, camera and triangulation information of the aerial photos. Digital elevation points produced by means of photogrammetric digitization of aerial photographs were transformed into grid model with grid interval of 0.5 meter. Changes took place on land surface were calculated by extracting digital elevation of 1995 from the digital elevation classes of 1969 which are produced by DTMs. Quantities of eroded and accumulated material were determined and expressed as t/pixel/26 years and t/ha/year by combining variation values, pixel areas and numbers and bulk density of soil samples. Calculation have revealed that 8 t/ha/year of soil has been lost including road construction in eroded areas, whereas 18.30 t/ha/year of soil has been gained in accumulation areas in various forms. By omiting road construction area's data, 13.08 t/ha/year of soil has been lost in eroded areas that includes erosion between 1 - 25 cm. Key Words : Erosion, gully, modelling, geographical information systems, remote sensing, Çubuk dam, image analysis INTRODUCTION The data accumulation studies realised for the purpose of modelling the soil erosion aims at the determination of the methods that can be used for the prevention of erosion and together with these methods can be adopted for the designation of the best model of land use. This can sometimes be in the form of defining the degree of erosion and sometimes in the form of constituting a model on the basin by having data about factors accumulated the factors that are controlling erosion. The high rate of resolution possessed by the remote sensing data and the developed data processing technologies have rendered possible the accumulation and evaluation of the partial data. The unwanted accumulation of the eroded material is a very general explanation for the sedimentation problem. At the basin areas, the problem regarding the general condition of sediment control initiates with the erosion of the soil by the effect of water or wind. Within the scope of the basin protection projects, the reduction of the amount of erosion and the improvement of sedimentation problems constitute the basic objectives. For this purpose, it is necessary to acquire information including the intensity of the problems encountered. This information that are necessary can be achieved within a short period of time with the help of Geographical Information Systems (GIS) and the Remote Sensing (RS) techniques. The GIS and RS techniques are gaining a significant importance in the recent years from the viewpoint of obtaining the data used in relation with the erosion and sediment estimation models and the model studies that are operated or formed at GIS environment with the adoption of these models. The data such as the cover of vegetation, the length of the slope, the characteristics of the soil, the hydraulic parameters and the land usage are providing numerous conveniences for the transactions regarding the transfer of the present maps to a digitised medium, the processing and storage of the acquired data, the obtaining of new data from the present data via cross - examinations and for the intensive and costly field studies lasting for long periods of time. This study has been carried out for the purpose of designating the change that has been formed by erosion from the environmental viewpoint at the areas surrounding the Çubuk Dam Lake by benefiting from the quantitative modelling for reasons as dictated above and to determine the dimensions of the erosion at this region. MATERIAL AND METHOD Material The investigation area is approximately 1000 hectares and covers the Çubuk 1 Dam Lake and the area surrounding it. The investigation area is situated at the North of Ankara and lies in the North - South direction. The distance of the Çubuk County that gives its name to the plain, to Ankara is 40 kilometres. The area lies between the 32o 68' - 32o 88' Western longitudes and the 40o 00' - 40o 03' Northern latitudes (Çinkaya 1993). In this study, the 1:21000 scaled aerial photographs for the year 1969 and the 1:4000 scaled aerial photographs for the year 1995 in relation to the area of study and obtained from the General Directorate of Title and Cadastre the General Directorate of Ankara Water and Sewer Administration, the elevation data obtained by the digitisation of these aerial photographs by the adoption of photogrammetric methods, the triangulation information regarding the aerial photographs and the camera calibration information, the satellite images, the geological, topographical and soil maps and the studies realised by other institutions regarding the subject have been used. The project has been carried out at the GIS and RS laboratories at the Soil and Fertilizer Research Institute and by benefiting from the softwares and hardware related with the geographical information systems of the Soil and Water Resources National Information Centre Soil and Water Resources Research Section of which is affiliated with the General Directorate of Rural Services. Within the scope of the study, TNTmips 6.2, Intergraph MicroStation 95, ArcInfo 7.2.1, Erdas Imagine 8.1 and ArcView 3.1 softwares have been used. On the other hand, the base materials used in the digitisation of the aerial photographs were the 1:25000 scaled topographic maps, 1:21000 scaled aerial photographs for the year 1969, 1:4000 scaled aerial photographs for the year 1995, the coordinates of the ground control points regarding this project and the camera calibration information. Method Within the scope of the study, the digital Terrain models obtained by using the elevation data acquired from the digitised aerial photographs have been used as a base. Taking horizontal cross - sections and the realisation of the volume analyses and erosion determination transactions on these bases constitute the method of the study (Welch et al. 1984, Dymond et al. 1986). In the preparation of the data that have been digitised by photogrammetric methods with the adoption of the Micro-Station 95 software, the stages involving the transfer of the aerial photographs to a digital medium through a scanning transaction, the realisation of Aerial Triangulation measurements by using the aerial photographs already transferred to a digital medium, the balancing of aerial triangulation, digitisation and the accumulation of DTM points have been followed. DISCUSSION Digital Terrain Models Within the scope of this study, grid models have been constituted from the digital Terrain models at 50 cm grid intervals, by using the digital elevation point data obtained by the digitisation of the aerial photographs at a scale of 1:21000 for the year 1969 and of the aerial photographs at a scale of 1:4000 for the year 1995 through photogrammetric methods. The grid models have been prepared by the use of the ArcView 3D modules. The digital elevation points adopted for the modelling studies are presented in Figures 1. and 2. On the other hand, the digital elevation classes produced from this digital Terrain model are presented in Figures 3. and 4. Since the IDW method has been adopted as a program interpolation technique, the grid system has also been realised by the adoption of this method.
Erosion Calculations The digital elevation classes of two different dates have been transferred into the ArcInfo program and their differences are taken. The new map thus obtained (Figure 5.) is currently representing the difference acquired as a result of erosion between the years in question. Depending on the positive and negative values of the pixels obtained from this map that has been formed and also on the constant areas, a classification has been realised and the results are presented in Figure 6. While the positive values in relation to the pixels stand for the presence of erosion, those of the negative values for the pixels stand for the presence of accumulation. By the studies performed on this map, it has been determined that the amount of erosion ranges between 1 cm and 60 cm and that the amount of accumulation ranges between 1 cm and 99 cm. The positive values were mostly obtained at the points with high steep slopes, at abondoned riverbeds and at the gully areas. At these areas, it was as well observed that there existed certain accumulation points and that especially at the upper sections of the area and at the regions where foresting has been realised, it was determined that the amount of accumulation exceeded that of the erosion. At the agricultural land located at the upper sections of the area of study, it was observed that there were no changes in general.
Within the scope of the study, the size of one pixel was organised so as to be at a value of 50 x 50 cm, while the area of the pixel was 0.25 m2. The difference in elevation against each of the pixel numbers is calculated in terms of cm and these values are multiplied by 0.25 m2, which is the area of a single pixel. Therefore, the values thus obtained were converted into m3. After this, the amount of soil lost at the pixel is converted into tons by multiplying the achieved value with 1.33 g/cm3 that represents the average of the specific gravity values found for the soil samples taken from the area of study. By multiplying the pixel numbers belonging to the elevation differences with the ton values, the amount of loss that was formed during a period of 26 years was determined. By the accumulation of these values, it was calculated that the amount of soil loss formed at the entire erosion pixels in a period of 26 years were 97326.19 tons. By the division of this value as calculated above to 26, it was found out that the amount of soil loss realised in a period of one year were 3743.315 tons. The total number of pixels for the eroded areas was determined as 8180267. By multiplying the pixel number with the pixel area, it was found out that the area of the land under the effect of erosion was 204.52 hectares, and that an erosion at a value of 8.00 tons/hectares/year was observed annually at the area subject to erosion, which is found by dividing the amount of soil lost annually from the pixels to the area of the pixels. Widespread soil group forming in at the area of study is the brown soil group at a rate of 94 %. Upon the comparison of the value of erosion obtained with that of the 10 tons/hectares/year erosion value that represents the value of the soil loss attained experimentally for the brown soil group in uniform parcel studies, it was observed that there exists a conformity between these values (Doğan and Küçükçakar 1996).
As can be observed from the erosion map, the area of erosion arising from the construction of roads that is regarded to be unnatural was subtracted from the total erosion values, and thus only the amounts of soil loss belonging to the natural areas of erosion were calculated. For this purpose, the pixel numbers of the areas that were subjected to erosion at an amount ranging between 1-25 cm were added and the total number of pixels were found out to be 7967843. By multiplying this number with 0.25 m2 that represents the area of a single pixel, it was found out that the area that was subjected to erosion at an amount ranging between 1-25 cm was 199.19 hectares. It was thus determined that within a total period of 26 years, a total amount of 67752.11 tons of soil was lost from these pixels and that an amount of soil at an annual rate of 13.08 tons/hectares/year is removed. The three - dimensional images of the area formed by imparting an exaggeration to the digital elevation data have been displayed (Figure 8.) and by also overlapping, the observation of the colour differences thus formed as well as the positive and negative changes has been rendered over the surfaces.
Cross - Sections These models of two different dates that were transferred to the Erdas Imagine program have been coincided on top of each other by the adoption of the later stack analysis. On the new layer thus obtained, horizontal cross - sections were taken at different points. The horizontal cross - section points are shown on the image in TIF format as presented in Figure 9. The report files of the horizontal cross - sections were transferred into the Excel program, the graphical representations of the elevation differences arising due to the erosion encountered at the area were drawn and these unscaled graphical drawings are presented through Figures 10 - 17. As can be observed from the examination of the cross - sections, various changes have been realised on the surface of the land during a period of 26 years depending on the characteristics of the topography. This abondoned river bed that was selected as one of the horizontal cross - sectional areas, as can be observed from the figure presented below (Figure 10 ), is one of the areas suffering from the greatest amount of erosion. Despite the abundance of erosion at the area, there exist scarce locations where there are no changes or where there are accumulation points, due to reasons such as the isolated covers of vegetation on the surface that reduces the erosion of soil or the presence of stones or coarse materials that are difficult to be moved by the surface flow. Since the area of study is not a basin that is bounded by the water separation line, even though it is not possible to monitor the changes at the upper and lower sections of the basin that would mean erosion, the following horizontal cross - sections have been taken for the purpose of examining the differences that have been formed from the viewpoint of erosion, ranging from the upper sections of the area towards the lower portions of the area where the slope is increasing. By taking cross - sections at the points of the area that can be defined as the upper, intermediate and the lower sections of the same area, the differences in the changes observed alongside a slope depending on the steepness of slope are presented in the sample horizontal cross - sections provided in figures 11, 12 and 13. At the areas where the above - presented horizontal cross - sections are taken, it was observed that the amount of accumulation, rather than the amount of erosion, was too much. The reasons for this are the foresting and terracing studies realised at the area and the fact that the location represents a accumulation point for the soil originating from the upper sections of the basin. Upon the examination of the cross - section 6 taken at the lowest section of the area, it was determined that the amount of erosion was much more than the amount of accumulation and that there exists an erosion despite the presence of a cover of vegetation at especially the locations where the steepness of the slopes are increasing towards the vertical. At the area of study, within the horizontal cross - section samples taken from the surface of the lake vertically towards the crest, it is clearly observed that the area of the lake is narrower at the upper sections of the lake due to sediments, while the surface of the lake is wider towards the crest; in addition it is observed especially in horizontal cross - section samples numbered (figures 14, 15) 5 and 6 that the soil coming via erosion after the year 1969 are accumulated at the upper sections of the lake and at the boundary area. Even though the degree of change within the horizontal cross - section samples taken parallel to the area of the lake seem to be very little, it is as well observed that there exists a certain amount of accumulation at the hillsides and a certain amount of erosion at the gullies between the hills ( Figures 16, 17 ).
The horizontal cross - sections are especially reflecting the change that has occurred at the base of the gullies in an excellent manner. The area left between the two horizontal cross - sections and the depth measurements to be realised at the field , as well as the volume of the soil moving from that point can be easily calculated; and besides, the progresses developed at the gullies can be calculated in a much less time period when compared to traditional field studies, by taking points with GPS at pre-determined time intervals and by adopting these data together with the others. RECOMMENDATIONS The quantitative modelling, which during the recent years presents a significant importance among the methodologies adopted for the designation of erosion and that has achieved major advancements in accordance with the improvements obtained in the computer technologies, is gaining an increasing importance. This investigation, which presents a difference from the viewpoint of the methodology approaches from the studies that have been realised in our country until today, is representing a first with this aspect. In this modelling study, the basic foundations of which are formed by aerial photographs, the frequent difficulties encountered regarding appropriate area and appropriate scale of photographing are arising due to the differences in the subjects of studying of the institutions responsible from the taking of aerial photographs in our country and the scale of the photographs in relation to these subjects, as well as the absence of the taking of photographs at regular intervals. In accordance with these problems, the scales of the aerial photographs that can be obtained are possessing different scales, while it was impossible to obtain appropriate photographs for past periods. Despite the difficulties dictated above, the fact that the recent advancements regarding photogrammetry in our country are being followed and that they are carried out successfully has rendered positive contributions to the study. By the help of the digitisation realised in accordance with the sensitivity of the study, the differences between the details of the scale were minimised. The area selected in accordance with the supply of photographs is rendering difficultly the studying of the entire parameters in the attempt to provide an explanation for the erosion and accumulation points. From the viewpoint of the better studying of the results of the erosion and accumulation, it is necessary to handle such studies on the basis of the basins bounded by the water separation line. Therefore, the source of the incoming soil can be more clearly explained and also the factors affecting accumulation as well as the erosion can be better interpreted. In order to be able to test the results of the studies, the execution of similar studies in the following years at the areas involving more frequent field observation values is gaining an increasing importance. In parallel with this statement, the increasing of the studies to be executed all over the country on the subject of constituting a database in our country shall render the acceleration of the studies on the basis of mostly databases such as the remote sensing and geographical information systems, as well as the increasing of the number of the related studies. This methodology that can generate reliable results within short period of time in the subjects such as the gully erosion that necessitate more comprehensive field studies, provide an opportunity for monitoring the changes that arise within short time intervals. REFERENCES Çinkaya, N. 1993. Ankara metropoliten alanı içerisinde kalan Çubuk vadisi ve çevresinin arazi kullanım planlaması. Yüksek Lisans tezi (basılmamış). Ankara Üniversitesi, Ankara. Doğan, O. ve Küçükçakar, N. 1996. Ankara yöresinde üniversal toprak kaybı eşitliği parametreleri. Köy Hizmetleri Ankara Araştırma Enstitüsü Yayınları, Genel Yayın No:199, Rapor Seri No:105, Ankara. Dymond, J. R. and Hicks, D. L. 1986. Steepland erosion measured from historical aerial photographs. Journal and Water Conservation. 41-4; 252-255. Welch, R., Jordan, T. R. and Thomas, A. W. 1984. A photogrammetric technique for measuring soil erosion. Journal of Soil and Water Conservation. 39; 191-194. |
