پديد آورندگان :
اميري،افشين دانشگاه تهران - دانشكده جغرافيا , عبدالهي كاكرودي، عطا الله دانشگاه تهران - دانشكده جغرافيا , قديمي، مهرنوش دانشگاه تهران - دانشكده جغرافيا
كليدواژه :
استخراج اتوماتيك خطواره , رادار , گسل دهشير , سنجش از دور
چكيده فارسي :
گسل به شكستگي هاي سنگ هاي پوسته زمين گفته مي شود كه سنگ هاي دو طرف شكستگي نسبت به همديگر حركت كرده باشند. اين جابجايي مي تواند از چند ميليمتر تا صدها متر باشد. انرژي آزاد شده به هنگام حركت سريع گسل هاي فعال عامل وقوع اغلب زمين لرزه ها است. گسل دهشير يكي از گسل هاي فعال در ايران مركزي است و با روند تقريبي شمال غرب- جنوب شرق و طول بيش از 350 كيلومتر يكي از بزرگترين گسل هاي فعال ايران مي باشد. نقشه خطواره ها در مطالعات گوناگون از جمله زمين شناسي، هيدروژئولوژي ، توپوگرافي و غيره نقش مهمي ايفا مي كند. امروزه با پيشرفت هاي گسترده در فنون سنجش از دور، بهتر مي توان خطواره ها را شناسايي نمود. هدف اين مطالعه بررسي قابليت داده هاي لندست 5، لندست7، لندست8، استر، سنتينل 1، آلوس پالسار و مدل رقومي ارتفاع در آشكارسازي خط گسل و بالا آمدگي سطوح پيرامون آن مي باشد. روش كار مبتني بر الگوريتم تشخيص لبه است. در اين تحقيق پارامتر هاي بهينه براي استخراج اتوماتيك خطواره براي منطقه مورد مطالعه تعيين شد. مقايسه و ارزيابي نتايج به دست آمده نشان داد كه براي آشكارسازي خط گسل، مدل رقومي ارتفاع با رزولوشن 5/12 متر با دقت 61/91 درصد و ضريب كاپاي 91/81 درصد بهترين نتيجه را داشته است. داده هاي سنتينل 1 نيز نسبت به تصاوير اپتيكال قابليت بيشتري در تشخيص خطواره هاي مرتبط با گسل داشته است.
چكيده لاتين :
Introduction
Remote sensing satellite data has been widely used as a source of information for geologists on a regional scale. Detecting lineaments by remote sensing in desert and semi-desert areas where bedrock is fully visible can provide better results. Two types of lineaments are usually distinguishable by remote sensing data, namely: 1. Positive lineaments, including ridge and dyke bumps, and 2. Negative lineaments consisting of seams, cracks and faults. The purpose of this study is to detect the lineaments associated with the Fault, one of the active faults in central Iran, using optical, radar and altimetry data.
Materials and Methods
By examining the different bands of Landsat 8 satellite in order to select the appropriate band for extraction of lineaments, it was concluded that the shorter wavelength due to more penetration and better interaction with the ground surface phenomena would be more accurate. As a result, bands with wavelengths close to the wavelengths of Band 2 are used. After the necessary preprocessing, the filtering operation was performed, local sigma filtering was applied to all images (Asher, Landsat 8, DEM 12.5m and SRTM 30 m). The local sigma filter uses the local standard deviation calculated for the filter box to determine valid pixels within the filter window. This filter replaces the pixel value with the average calculated from valid pixels inside the filter box. In addition, Li filter were applied on radar images (Sentinel 1and Alos Palsar). In the present study, the automatic extraction of lineaments is based on two main calculations: first, the use of filters to detect edges, second, the information that gives us sudden changes in the value of neighboring pixels. Usually it is related to lineaments. The second stage reveals the lineaments
Results and discussion
In general, for fault detection, radar images are better than optical images. The DEM 12.5 m had the best accuracy among the other data sets. Among the optical images, Landsat 8 OLI sensor data with 30 m spatial resolution was more capable of fault detection. Sentinel-1 images in C band is more capable than Alos palsar L-band radar images. In the northern sections of the fault, the eastern plate of the Dehshir Fault, show an uplift. In the southern part of the fault the western plate of the fault is uplifted. The Dehshir fault moves in both horizontal and vertical directions.
Conclusion
In this study, using the remote sensing data (optical, radar and digital elevation model), the Dehshir Fault, which is an active strike-slip fault, is detected. Remote sensing data are particularly important in radar extraction for geological and geomorphological applications. Radar data have been able to identify fault lines in almost all parts of the area due to their better interaction with surface phenomena. Optical data is not well capable as radar images for extracting fault line. By combining remote sensing techniques with fieldwork, you can achieve desirable results with lower cost and better accuracy.
