شناسایی و استخراج تغییرات بستر ماسوله رودخان استان گیلان در بازه زمانی 2000 تا 2020 با استفاده از پردازش تصاویر ماهواره ای

Abstract:
As a dynamic system, the river always changes its location and morphological characteristics according to time, geomorphic, geological, hydrological factors and sometimes due to human intervention. Masuleh Rudkhan is one of the most important rivers in Iran that the formation of bed geometry in different periods is very different from each other. This river has a dynamic morphological behavior under the influence of various factors such as the geology of the region, the characteristics of the alluvial structure, the hydrological characteristics of its upstream basin, the structures in it and the hydraulic conditions of the flow. The purpose of this paper was to identify and extract the changes of Masouleh Rudkhan in Guilan province in the period 2000 to 2020 using satellite image processing. The images used were the images of Landsat 5 satellite on 06/06/2000 and Landsat 8 on 13/06/2020. SVM classification method and NDWI, MNDWI, AWEI and WRI indices were used for image processing. Changes in the SVM classification method showed that the river area has decreased by 314.26 hectares from 2000 to 2020. These changes mean an increase in construction on the riverbed and a decrease in the amount of water in the river. Comparing kappa coefficient and overall image processing accuracy, it was observed that AWEI index with kappa coefficient and overall accuracy of 0.93 and 0.95 in 2000 and kappa coefficient and overall accuracy of 0.94 and 0.96 in 2020 had the highest accuracy and Masouleh Rudkhan route in this index to Google earth moved. By examining the river route in a period of 20 years, it was observed that at 5 km from the beginning of the basin, the river route is 100 meters to the south, at 7.5 km, the river route is 50 meters to the south, at 29.3 km, it is 45 meters to On the south side, at km 48, it is 38 meters to the north and at km 50 to 56, it has changed continuously and the reason for this change is the high erosion in this part of Masouleh River.
Keywords: Masuleh River, changes, water index, remote sensing

1- Introduction
The emergence and formation of natural waterways, which are the same as rivers, is affected by various factors such as climatic conditions, geology and geography. A river may form in an erosion-resistant structure, in which case its position and dimensions remain constant. On the other hand, the river can appear in eroded alluvial materials. In such a case, a constant tendency to continuously change the position of the river through erosion and reformation of coastal walls along the sides is clearly visible. Changes in the geometric pattern of the river, bed instability and erosion of the river banks not only cause damage to agricultural lands and facilities adjacent to the river bank, but also sediments from erosion reduce the useful capacity of dams and disrupt the hydrological conditions of the flow. Rivers rarely have regular and direct flow all the way. Rather, depending on the structure of the land, the shape of the roughness and also the climatic conditions, the dynamics of the river flow is unstable and causes changes in the waterways and twists the course of the river (Ahmadi, 2006). Monitoring the morphological changes of the riverbeds is essential and in the meantime, remote sensing is a powerful tool. With this approach, in this study, identification and extraction of the changes in Masouleh Rudkhan of Gilan in the period of 2000 to 2020 has been conducted using satellite image processing.

2- Materials and Methods
• Research Method
The method of conducting research was spatial-field analysis; in this way, after conducting preliminary studies and preparing appropriate satellite images, the necessary processing was performed. The images used in the research were Landsat 5 satellite images dated 2009/06/06 and Landsat 8 images dated 2020/06/13 which are illustrated in Table (1) of the specifications of the images used. In order to study the changes in the river route, images were classified and indicators were applied, which are described in detail.
• Pre-processing the Images
1. Geometric Corrections: Usually digital images have significant deviations, so that such images cannot be used directly as surface maps. Geometric corrections are meant to compensate for errors so that the corrected digital layer is able to adapt to ground surface positions (Shenani Hoveizeh and Zarei, 2016). In this study, the vector layer of waterways and route as well as 200 land points were used.
2. Atmosphere Corrections: Atmosphere corrections are also necessary in cases where the intensity of the signal sent by phenomena such as the ocean is less than the atmospheric effects. Water absorbs electromagnetic waves, especially in long wavelength bands such as red and infrared; therefore, the amount of reflected energy from the surface of water zones is much less and most atmospheric effects are received and recorded by satellite sensors (Fatemi and Rezaei, 2012). In this study, the numerical value reduction method of dark pixels has been utilized for atmospheric correction of images. In this method, a constant value of the total pixels in a given band is reduced to apply atmospheric corrections to each satellite image (Chavez, 1988).

Discussion and Conclusion
Masouleh Rudkhan is one of the most important rivers in Iran that the formation of bed geometry in different periods is very different from each other. This river has dynamic morphological behavior under the influence of various factors such as geology of the region, characteristics of alluvial formations, hydrological characteristics of its upstream basin, structures in it and hydraulic conditions of the stream. The purpose of this article is to identify and extract the changes of Masouleh Rudkhan in Gilan province in the period 2000 to 2020 using satellite image processing. The images employed were the images of Landsat 5 satellite on 2009/06/06 and Landsat 8 on 2020/06/13. SVM classification method and AWEI, MNDWI, NDW, and WRI indices were used for image processing. The changes studied in the SVM classification method showed that the river area has decreased 26,314 hectares from 2000 to 2020. These changes mean an increase in construction on the riverbed and a decrease in the amount of water in the river. Comparing Kappa coefficient and overall image processing accuracy, it was observed that AWEI index with Kappa coefficient and overall accuracy of 0.93 and 0.95 in 2000, and Kappa coefficient and overall accuracy of 0.94 and 0.96 in 2020 had the highest accuracy and Masouleh Rudkhan route was transferred to Google Earth in this index. Examining the river route over a period of 20 years, it was observed that at 5 kms from the beginning of the basin, the river route is 100 meters to the south, at 7.5 kms, the river route is 50 meters to the south, at 48 kms, as much as 38 meters to the north, and at 50 to 56 kms, it has changed direction continuously, and the reason for this change of direction is the high erosion in this part of the Masouleh river.