Evaluation of Active Geodynamics via Interfering Techniques in SAR Images (Case Study: Bardsir Plain)
محورهای موضوعی : فصلنامه علمی پژوهشی سنجش از دور راداری و نوری و سیستم اطلاعات جغرافیایی
1 - M. A. Student of Remote Sensing and Geographic Information System, Yazd Branch, Islamic Azad University, Yazd, Iran
کلید واژه: Filter, Subsidence, Sentinel 1, Bardsir plain, radar interferometry technique,
چکیده مقاله :
In recent years, it has occurred in different regions of Iran, especially the plains, and in most regions has caused this phenomenon to become a major regional and country crisis. Kerman desert province is no exception to this rule and most of its plains and industrial areas have suffered from this phenomenon and have high subsidence rates. The present study investigated the occurrence of this phenomenon using radar interferometry technique and Sentinel 1 satellite images in the period of 2019 and 2020 in Bardsir plain of Kerman province. To investigate the rate of subsidence in the region, initial processing was performed in remote sensing software and GIS and two Goldstein and Adaptive filters were used to evaluate the obtained results. The results show that the Goldstein filter has subsidence values up to 10 cm in certain ranges and the uplift values up to about 6.5 cm and the Adaptive filter have given the subsidence values up to 9 cm in some ranges and the uplift values up to about 5.6 cm. The reason for the difference in values in the results of these two filters is that in the Goldstein filter, the amount of coherence increases by manipulating the phases, so the image is brighter, thus the situation in this filter improves. But this is not the case with the Adaptive filter, and the phases are not manipulated, and in some areas, the amount of blurriness is higher in different parts of the image.
In recent years, it has occurred in different regions of Iran, especially the plains, and in most regions has caused this phenomenon to become a major regional and country crisis. Kerman desert province is no exception to this rule and most of its plains and industrial areas have suffered from this phenomenon and have high subsidence rates. The present study investigated the occurrence of this phenomenon using radar interferometry technique and Sentinel 1 satellite images in the period of 2019 and 2020 in Bardsir plain of Kerman province. To investigate the rate of subsidence in the region, initial processing was performed in remote sensing software and GIS and two Goldstein and Adaptive filters were used to evaluate the obtained results. The results show that the Goldstein filter has subsidence values up to 10 cm in certain ranges and the uplift values up to about 6.5 cm and the Adaptive filter have given the subsidence values up to 9 cm in some ranges and the uplift values up to about 5.6 cm. The reason for the difference in values in the results of these two filters is that in the Goldstein filter, the amount of coherence increases by manipulating the phases, so the image is brighter, thus the situation in this filter improves. But this is not the case with the Adaptive filter, and the phases are not manipulated, and in some areas, the amount of blurriness is higher in different parts of the image.
Evaluation of Active Geodynamics via Interfering Techniques in SAR Images
(Case study: Bardsir plain)
Zohre Hamzeh
M.A student of Remote Sensing and Geographic Information System, Yazd Branch, Islamic Azad University, Yazd, Iran.
Abstract
In recent years, it has occurred in different regions of Iran, especially the plains, and in most regions has caused this phenomenon to become a major regional and country crisis. Kerman desert province is no exception to this rule and most of its plains and industrial areas have suffered from this phenomenon and have high subsidence rates. The present study investigated the occurrence of this phenomenon using radar interferometry technique and Sentinel 1 satellite images in the period of 2019 and 2020 in Bardsir plain of Kerman province. To investigate the rate of subsidence in the region, initial processing was performed in remote sensing software and GIS and two Goldstein and Adaptive filters were used to evaluate the obtained results. The results show that the Goldstein filter has subsidence values up to 10 cm in certain ranges and the uplift values up to about 6.5 cm and the Adaptive filter have given the subsidence values up to 9 cm in some ranges and the uplift values up to about 5.6 cm. The reason for the difference in values in the results of these two filters is that in the Goldstein filter, the amount of coherence increases by manipulating the phases, so the image is brighter, thus the situation in this filter improves. But this is not the case with the Adaptive filter, and the phases are not manipulated, and in some areas, the amount of blurriness is higher in different parts of the image.
Keywords: Subsidence, Bardsir plain, Radar interferometry technique, Sentinel 1, Filter
Introduction
The Earth's crust is not always calm, but under the influence of tectonic factors, it gradually moves up and down, which can be seen in the discontinuity, deformation, and cessation of sedimentation, as well as uplift in different parts of the earth [Ahmadi et al., 2016]. Land subsidence is one of the natural hazards that often occurs vertically and is not felt in a short time [Afifi, 2017]. This phenomenon is a global problem and more than 60 countries around the world are facing issues related to this problem. Land subsidence usually leads to damage to the aquifer system, reduced water quality, and the destruction of underground and surface structures such as underground tunnels, buildings, roads, railways, and pipelines [Xi-Cun et al., 2019].
Increasing groundwater exploitation, especially in basins that have been accumulated by alluvial deposits, shallow seas, or unconsolidated lakes, can cause subsidence or subsidence. According to the United States Institute of Geology, the phenomenon of subsidence includes collapse or subsidence that can have a slight horizontal displacement vector [Sharifi Kia, 2013]. Many plains around the world are affected by land subsidence, which often includes various geodynamic, hydrogeological, climatic, and social fields [Li et al., 2019]. But Interferometric Synthetic Aperture Radar or INSAR is a telemetry technology invented to study the surface motion of the ground at the 1992 Landers, California earthquake. In the INSAR technique, motion, altitude, and ground surface changes are mapped using a combination of data obtained from radar imaging systems (SARs) mounted on satellite or aircraft platforms [Nedai'e, 2017]. Over the past decades, time series radar interferometer (T-InSAR) has emerged as a powerful technique for measuring various surface deformation phenomena [Samiei Esfahany et al, 2019]. The basis of work in measuring ground surface movements is the use of repetitive radar images. An image captured from one area at a given time (reference time) is combined with an image captured at the same time by the same radar sensor (following time). With this technology, movements and changes caused by phenomena such as earthquakes, volcanoes, glaciers, landslides and diapirism of salt domes or irregular phenomena such as groundwater and oil outflows, farm irrigation and underground explosions can be studied. In addition, surface phenomena such as fires, floods, changes in humidity and plant growth can be detected. Therefore, its applications include the study and identification of natural and human hazards and the quantification of human relations with natural resources [Nedai'e, 2017]. Due to the importance of this issue, a lot of research has been done in the world and Iran. Goli Mokhtari et al. (2017) evaluated the subsidence of Nurabad plain aquifer subsidence using the radar interferometry method. The results showed that in the plain area in a period of 12 years, the highest rate of groundwater loss was about -22 meters in the eastern, southern, and central parts of the aquifer located in the piezometric wells of nomadic settlements, Khomezar, industrial town, Tale Meshki, and Shur. Subsidence maps showed that the average settling rate was 4 cm per year over a 4-year period of about 0.16 m (16 cm) [Goli Mokhtari et al., 2020]. Ghahroudi Tali et al. (2017) studied the evaluation of slope instability in the Lorestan railway area using the radar differential interferometry (DInSAR) method and the results of this study showed that radar data and differential interferometric processing method due to wide coverage and frequency high data accuracy has the potential to detect the instability of slopes and calculate their displacement. Interpretation of time series images showed that the highest amount of amplitude material movement occurred in autumn and spring and the highest amount of amplitude material movement in the period 2015 to 2018 is about 28.8 cm in the range of Tang Haft to Tang Panj station, which indicates the activity of the region in terms of movements [Qahroodi tali et al., 2020]. Keshavarz et al. (2015) studied the subsidence of Faryab Industrial Zone using Sentinel 1 images for 2015-2017 and the results showed that the average land subsidence in 2015 was 7 cm, in 2016, 9 cm and in 2017, it has been 7 centimeters [Keshavarz et al., 2019]. Bayati et al. (2015) in a study entitled "Comparative evaluation of methods for measuring morphotectonism by analyzing satellite images in Zayandehrud and Abdughi areas studied the technical and morphotectonic situation and made a comparison between morphological indices and the results of radar data in these areas." The results showed that in both regions, the displacement values show a very high correlation and coordination with the results obtained from morphic indices and there is a small amount of coordination only in a few points [Bayati Eshkafti et al., 2019]. Imam Bakhsh et al. (2015), in a study, studied land subsidence in Arzooieh plain of Kerman using radar differential interferometry (DINSAR) method and using ASAR and Sentinel 1 sensors. The results of the analysis of the obtained time series showed that the region is continuously subsiding; The amount of subsidence from 2008 to 2009 is equal to 15 cm, which is the most subsidence in the northwestern and central part of the plain, and from 2009 to 2010 it is 11.9 cm, which is again the highest amount of subsidence in the northwestern and central part of the plain and, for 2014 to 2015 is equal to 12.8 cm and the highest amount of this phenomenon is in the southeastern part of Arzooieh plain. Examining the subsidence images of these years, it can be concluded that the subsidence of Arzooieh plain has a trend from northwest to southeast of the plain [Imam Bakhsh et al., 2018]. Piroozi (2017), in a study that analyzed land subsidence in the plains around Tehran, reviewed the site and collected records. In this article, the change of land level in Tehran and Shahriyar plains during the years 1992-2010 and its relationship with changes in underground height in this period has been investigated. The results showed that in some areas, other factors such as underground erosion, and the presence of old aqueducts, drainage of organic soils, and soil removal were more problematic than groundwater depletion [Pirouzi and Eslami, 2019]. Ji Kan Hee (2019) examined land subsidence control and environmental protection policy in Shanghai. The results of the study showed that to control subsidence, Shanghai identified three areas of subsidence control, in which special proceedings have been taken. In terms of Strategic Environmental Assessment (SEA), the 2013 regulation scores high, indicating that groundwater abstraction and recharge control is effective. Landslides observed over the past six years also confirm the effectiveness of the 2013 regulation with the SEA's strongest concern for the protection of the sustainable environment in Shanghai [Xi-Cun et al., 2019]. Salehi Motahed et al. (2019), in their research, examined the effective geological parameters on land subsidence in Mashhad plain, northeastern Iran. The results showed that the assessment of geological features of the subsidence area, in addition to groundwater depletion, tectonic status, and active faults that cause changes in bed depth, play a key role in the pattern and extent of subsidence. Also, stabilization of Pleistocene marl bed located in the central part of the plain was effective in increasing the subsidence of the study area [Salehi Moteahd et al., 2019]. Sedaghat et al. (2020) studied the assessment of subsidence risk due to groundwater abstraction in Isfahan metropolis and the results showed that by moving from southeast to northwest in the study area, groundwater reduction from 4 to 36 meters was increased. According to the obtained results, the most severe subsidence occurred in the western border of the city and has been calculated to be about 73 cm during 14 years. Besides, the specific non-static skeletal reserve (Sskv) for fine-grained sediments was calculated during the mentioned period, and based on the interpolation map, the maximum specific values of rigid skeletal storage, as well as the maximum, predicted subsidence due to gradual depletion of groundwater had happened in central parts of the city [Sedaghat et al., 2020]. The purpose of this study is to obtain subsidence in Bardsir region in Kerman province using radar interferometry technique and Sentinel-1 images in 2019 and 2020. Given that no research has been done on the issue of subsidence in this area, so this issue is new and up to date.
Research Method
Study Area
Bardsir city is one of the cities of Kerman province which is located in the center of this province [Ahmadi et al., 2019]. This city is bounded by longitudes of 56°30 'and 57°.00 east and latitudes of 29°30' and 30°.00 north, is located in the southwest of Kerman. The city of Bardsir is located in the northwest of this square. The area has two mountain ranges, one of which is located in the northwest-southeast trend in the north and the other with the same trend, covers the southern half.
Figure 1. Study map of the area
In this study, Sentinel-1 radar satellite images in 2019 and 2020 have been used to investigate the amount of subsidence in the area. These images were downloaded and used from the Sentinel database. In Tables (1) and (2), the information about the captured images and the orbit of the related files can be seen. The Sentinel 1 captures ground-based radar in the C-band and provides imaging in four different styles or modes with acceptable resolution up to 5 meters and coverage within 400 kilometers. Its polarity orbit and the dual-polarization capability of this satellite have led to the rapid delivery of data from the satellite to ground stations. The satellite has virtual aperture radar that provides high-resolution images. The satellite is also designed to capture uninterrupted imaging of all areas of the planet, including coastal areas, transportation routes, and large-scale ocean cover in uninterrupted operation [Keshavarz, 2019].
Table 1. Sentinel-1 image information
Satellite direction of movement | Polarization | Date | Image type | Image mode |
Down pass | VV VH | 2019.02.09 | SLC | IW |
Down pass | VV VH | 2020.02.04 | SLC | IW |
Table 2. Orbit Information OF Sentinel-1 Images
FILE NAME | IMAGE DATE |
S1A_OPER_AUX_POEORB_OPOD_20190301T121016_V20190208T225942_20190210T005942 | 2019.02.09 |
S1A_OPER_AUX_POEORB_OPOD_20200224T120915_V20200203T225942_20200205T005942 | 2020.02.04 |
Workflow diagram
Figure 2. Research flowchart diagram
Radar differential interferometry technique
In radar interferometry, the resulting phase of two images taken from a given area is interfered with to produce an interferometer. In fact, the interferometer is the product of the mixed product of two radar images. These two images may be taken by an aerial or space platform that has two antennas at a certain distance (baseline) (interference measurement with a single passage) or two images with different time intervals and taken from the same platform (interference method with frequent crossings). The phase difference in the two images is shown as an edge in the interferometer, where each edge is associated with a 2π phase difference. The interferometer provided by the InSAR method is capable of displaying height changes and roughness. The accuracy of changes in altitude values (dz) can be calculated from any margin as a function of satellite characteristics such as baseline length (Bn), wavelength or band used (l), angle of impact (q) and mile length from satellite height to ground (p) (dz = (lp Sinq) / 2Bn). The radar image with virtual apertures consists of the phase wave amplitude of the return of the radar signal. Based on research and observations, the recorded phase (ΔΦ) of the wave contains better information and properties than the amplitude of the wave (Δp) in radar images [Ghiglia and Pritt, 1998]. According to this calculation, the height of a point obtained from its amplitude difference is (Δr = p-p!) possible by calculating the phase difference expressed (ΔΦ) in the interferometer and with the help of the following equation:
(1)
|
|
The radar interferometry (InSAR) method makes it possible to produce digital models of terrestrial roughness with optimum elevation accuracy for 5.6 cm C-band data with a wavelength of about 5 m [Sharifikia, 2006]. This method is able to detect surface changes occurring on the ground at different intervals with millimeter accuracy using at least three (two DEM + images) or more radar images. The first image sensor records this space at time t0 and measures its phase value. Fm is the amount of subsidence of the distance P to P1 that has taken place during a certain time (Dt). To measure this value, the second image sensor takes a time at t with a geometry quite similar to the first image and measures the phase value on it. Differential sensing interference method, displays the phase difference Fs and Fm in the form of phase interference and graphs (DFint). If the surface is stable, the phase difference between the two images (SP-MP) is due to the change in position of the two sensors and its value is obtained by the following equation:
