آشکارسازی لندفرم های کلان حوضه یزد – اردکان ( با رویکرد کمّی) (مقاله علمی وزارت علوم)
درجه علمی: نشریه علمی (وزارت علوم)
آرشیو
چکیده
شناسایی لندفرم ها از مطالعات پایه ای در دانش ژئومورفولوژی است. اهمیّت شناسایی لندفرم ها به علت کاربرد آن ها در انواع برنامه ریزی های روستایی و شهری، برنامه ریزی آمایش و توریسم است. روش های چشمی در انتخاب بهترین ترکیب های باندی برای شناسایی لندفرم ها هم وقت گیر است و هم به علت ذهنیت گرایی و اعمال سلیقه های شخصی از دقت کافی در تشخیص حدود مرزی لندفرم ها و گاهاً نوع لندفرم ها برخوردار نیست. حل این مسئله از طریق کنکاش های رقومی در متن تصاویر قابل بررسی است. در این تحقیق از روش ترکیب آماری برای معرفی انواع حالات مختلف ترکیبات باندهای انعکاسی و از روش شاخص ترکیب بهینه باندی جهت انتخاب بهترین ترکیب باندی باهدف آشکارسازی لندفرم های کلان حوضه یزد اردکان در متن تصاویر سنجنده ETM+لندست از سری نسل هفتم استفاده شده است. نتایج حاصل از به کارگیری شاخص بهینه باندی در آشکارسازی لندفرم های کلان حوضه یزد اردکان نشان داده است که بهترین ترکیب باندی از بین بیست ترکیب مختلف باندهای طیفی سنجنده ETM+، ترکیب دو سه چهار با مقدار آماری 54.01 و نیز ترکیب یک دو چهار با مقدار آماری 54.02 است، به طوری که انواع دشت سرهای پخش سیلاب، اراضی مرتفع، اراضی کم ارتفاع و کویر یا شوره زارها و همچنین پدیمنت ها شناسایی شده اند.Detection of the macro landforms of Yazd-Ardakan basin (a quantitative approach)
Introduction Identification of landforms is one of the basic studies in the Knowledge of geomorphology. The importance of identifying landforms are due to their application in various rural and urban planning, landscaping and tourism planning.Traditional methods of identifying landforms like to Visual interpretation, are both time-consuming, boring and costly, particularly on a large scale and are not sufficiently accurate in identifying the boundaries of macro-landforms. Subjectivity and the application of personal tastes is one of the most important issues in only traditional methods. Because it does not have enough accuracy in determining the boundaries of landforms and sometimes the type of landforms. The solution of this problem can be investigated in digital images context analysis. In this research the study area is Yazd-Ardakan basin.It is located almost in the center of Iran. This basin is geomorphologically limited to Ardakan Playa from the north and from the south direction is limited to Shirkuh heights, and from the east to Khoranagh sub-basin and finally from the west to Nodooshan sub-basin is connected. The purpose of this study is Detection of the macro geomorphological landforms in Yazd Ardakan basin using optimum composition index. Materials and methods In this research, the ETM+ sensor data from 7th generation Landsat satellite has been used to identify the large landforms such as Playa, pediments and, etc. The data of this sensor have many applications in the field of geomorphology such as shoreline displacement, sedimentary and erosion basin identification, mountain front identification, plains, beaches, separation of lithological structures, water networking and, etc. which in this study are geomorphological Yazd-Ardakan has been discussed. In this study, were used 9 bands of ETM + data related to 162 pass and 8th row and dated 19 August 2010, which were almost without cloud cover or cloud free. Exact time of imaging according to the information in the metadata file was nine hours, eleven minutes and thirty-nine seconds. Also, the general geomorphology map extracted from the geomorphology map of Iran related to the Institute of Geography of the University of Tehran and also Google Earth software have been used to investigate the accuracy and compatibility of the identified landforms resulting from digital processing of satellite images with terrestrial reality. In this study Geo-statistical methods were used.The Optimum Index Factor provides the best and most suitable band composition among the possible types of spectral band compositions based on the amount of total variance and correlation between band compositions.In this method, the selection of the optimal combination between the bands is done by quantitative evaluation of the effects in the image.This method avoids wasting time due to the large number of possible RGB compounds in the visual analysis process. OIF values are used to determine the most optimum bands composition, and bands are ranked according to their statistical information, which includes total variance and correlation between different bands.The best band combination of all three possible band combinations has the highest amount of information and the lowest repetition rate. In this study, the "law of composition" in statistics was used to select "r" objects (here a combination of 3 bands) from "n" objects (here 30-meters Landsat bands). This law is in the following relation: C(n,r)=n!/r!(n-r)! Where: C(n,r) is a combination of r object from n object and sign! It is called factorial and the factorial of any natural number (here 6 spectral bands of ETM + sensor) means the product of that number multiplied by all natural and integer numbers before itself In general, the product of n × (n-1) × (n-2) ×… 2 × 1 Therefore, 20 bands compositions can be written for ETM+ sensor reflective bands, of which only one combination is the most optimal and desirable band composition for displaying geomorphological features of the earth surface due to having the most non-repetitive information, and it is the compound that have with the highest OIF statistical value. T In order to calculate the OIF that shows better the surface landforms on a large scale, the correlation coefficients between the different bands must be calculated. Here, to show the degree of correlation between different ETM+ bands, a correlation matrix is used, which have always number one in his main diagonal. Result and discussion First, the basic statistics (including mean, minimum, maximum, standard deviation and eigenvalues) were calculated for 30-meter bands of ETM+ sensor. The mentioned parameters are used in calculating OIF. Then the correlation matrix was calculated between ETM+ spectral bands in the ENVI software. Finally, the OIF index table was obtained for all possible three-band combinations by the basic statistics table and the correlation matrix table. The results of the recent table showed that the best and most optimal band compositions that lead to the maximum Detection of landforms are related to rows 12 and 20 of this table, because the mentioned rows had maximum OIF values. Other results of this research were the produce of a raster map and vector map of identified large landforms obtained by OIF. Conclusion This study showed that the higher the standard deviation between a three bands combination and the correlation coefficient between them be less, the OIF index have the higher value. This means that the three bands color combination is one of the most non-repetitive information items. In this study, two bands combinations of 124 and 234 of the ETM+ sensor had the highest values, and both combinations had approximately 54 OIF values. The advantage of calculating this index are 1. Finding the best three bands color combination to highlight and Detection the image, 2. It is fast and repeatable due to digital image processing so that if we want to visually determine the best three bands combination, it is very boring and time-consuming. The Results of the OIF index has shown that the best bands of twenty different combinations, is two-three-four spectral combination. Because the OIF index was greater than the other band combinations (OIF=54.01). The combination of recent spectral bands (two-three-four) led to identify the mountainous and flood spreading pediments too Sebkha landform.
تبلیغات
