بررسی کیفیت مدل رقومی ارتفاعی تهیه شده از تصاویر سنتینل -1 برای استخراج آبراهه ها: مطالعه تطبیقی در حوضه آبخیز تفتان (مقاله علمی وزارت علوم)
درجه علمی: نشریه علمی (وزارت علوم)
آرشیو
چکیده
بررسی مورفومتری یک حوضه آبخیز نقش بسیار مهمی در ساختار زمین دارد. روش های سنجش از دور ابزار مناسبی برای بررسی و استخراج شبکه آبراهه به شمار می آیند. یکی از روش های رایج برای استخراج شبکه آبراهه و بررسی مورفومتری حوضه آبخیز بهره گیری از مدل های رقومی ارتفاعی با قدرت تفکیک مکانی زیاد است. هدف از این پژوهش، استخراج شبکه آبراهه با استفاده از مدل های رقومی با قدرت تفکیک مکانی زیاد مانند آلوس و سنتینل -1 و مدل های رقومی با قدرت تفکیک مکانی متوسط مانند اس آر تی ام و تان دم ایکس است. به منظور تهیه مدل رقومی ارتفاعی سنتینل -1 از روش تداخل سنجی راداری استفاده شد. درنهایت، برای اعتبارسنجی دقت این مدل رقومی ارتفاعی از آلوس استفاده شد. نتایج نشان دهنده آن است که مدل رقومی ارتفاعی تهیه شده از سنتینل -1 با داده مرجع در این پژوهش (آلوس) دارای همبستگی حدود 99/0 که نشان دهنده قابلیت زیاد این مدل رقومی ارتفاعی در استخراج شبکه آبراهه است. نتایج استخراج آبراهه حاکی از آن است که دو مدل رقومی ارتفاعی با قدرت تفکیک مکانی زیاد سنتینل -1 و آلوس هرکدام تعداد 9 شبکه آبراهه را استخراج کردند؛ در حالی که مدل های رقومی با قدرت تفکیک مکانی متوسط اس آر تی ام و تان دم ایکس به ترتیب توانستند فقط 7 و 6 شبکه آبراهه را استخراج کنند. بررسی ها در این پژوهش نشان دهنده آن است که برای بهبود کیفیت مدل رقومی ارتفاعی سنتینل -1 باید پارامترهای خط مبنا و همچنین اختلاف زمانی بین دو تصویر قدیم و جدید در تداخل سنجی راداری بسیار موردتوجه پژوهشگران قرار گیرد.Evaluation of the Quality of a Digital Elevation Model (DEM) Prepared by Using Sentinel-1 Images for Extracting Waterway Networks: A Comparative Study in Taftan Watershed
Morphometric study of a watershed plays a very important role in land structure. Remote sensing methods provide a good tool for studying and extracting stream networks. One of the common methods for extracting waterway networks and conducting a morphometric study of watersheds is the use of Digital Elevation Models (DEMs) with a high spatial resolution. The purpose of this study was to extract stream networks by using the DEMs of high spatial resolution, such as ALOS-1 and Sentinel-1, and those of medium spatial resolution like SRTM and TanDemX. To produce the DEM by using Sentinel-1 images, the InSAR method was applied. Finally, to validate the accuracy of this DEM for checking Sentinel-1 ability to extract stream networks, the ALOS-1 DEM with the spatial resolution of 12.5 m was used. The results revealed that the produced DEMs by using Sentinel-1 images had a high correlation of about 0.99 with the reference data of ALOS-1, thus showing the high capability of the DEM for extracting stream networks. The results of extracting the waterway networks demonstrated that each of the two DEMs of Sentinel-1 and ALOS-1 with high spatial resolutions could extract 9 waterway networks, while the digital models of SRTM and TanDemX with medium resolutions could only extract 7 and 6 stream networks, respectively. The studies indicated that the baseline parameters, as well as the time difference between the two Master and Slave images in InSAR, had to be highly considered by researchers to improve the quality of the Sentinel-1 DEM.Keywords: Digital Elvation Model (DEM), Sentinel-1 Image, InSAR, Stream Extraction, Remote Sensing, Taftan IntroductionStream network is an important feature for hydrological modeling, geomorphological analysis of the landscape, and many other applications. Recently, Digital Elevation Models (DEM) have emerged as a powerful tool for assessing landscape deformation via time, place, and especially stream network analysis, which itself requires a DEM of high spatial resolution. In Iran, due to the lack of DEMs with high spatial resolutions, more digitized topographic maps are used, which are time-consuming to scan. With recent advances in optical and radar remote sensing, these limitations can be partially overcome and various methods can be used to produce DEMs. The main advantage of using radar data compared to other methods is that these types of satellites can capture images of the regions in any weather conditions and even during day and night. This can be an acceptable method for producing DEMs with high spatial resolutions in different spatial and temporal conditions. One of the most important and common methods of producing the DEMs of radar data is using InSAR technique, which considers the phase difference between two radar images to produce DEMs. So far, few studies have been conducted to produce a DEM with high spatial resolution and finally accurately extract waterway networks in Iran. Therefore, the aim of this study was to generate a DEM between Sentinel-1 images by using InSAR and evaluating it with the ALOS-PALSAR DEM. Finally, using the sentinel-1, ALOS, TanDemX, and SRTM DEMs, the stream networks could be extracted in the study area. Materials and MethodsIn this study, two images of Sentinel-1 with the time difference of about 12 days and baseline of 161 were used by InSAR to produce a DEM. The D8 algorithm was used to extract the stream networks from the DEMs of ALOS-1, SRTM-1arc, TanDemX, and Sentinel-1 in the study area. Research findingsThe DEM produced by Sentinel-1 had a correlation of 0.99 with the base-altitude digital model (ALOS), but the standard deviation between these two data was 31 m despite the obtained positive result. The DEMs of Sentinel-1 and ALOS-1 were both able to extract 9 rates of stream networks, but the DEMs of SRTM and TanDEMX were both able to extract 7 and 6 stream networks, respectively. Discussion of Results & Conclusion Analysis of the results obtained from the compatibility of the DEMs of Sentinel-1 and ALOS-1 showed that the high correlation between these two data could confirm extractions of the stream networks, which exactly resulted in the same extraction of the stream rate. On the other hand, SRTM and TanDemX satellites both had relatively similar results, but the reason for not extracting a rate with TanDemX compared to SRTM was the different spatial resolutions between the two satellites.
تبلیغات
