Iranian Jornal of Watershed Management Science&Engineering

fa تحلیلی بر عملکرد ماهواره آلوس جهت شبیه سازی سیلاب در دو حوضه با ویژگی‌های متفاوت Performance Evaluation of the ALOS Satellite for Flood Simulation in Two Hydrological Basins with Contrasting Characteristics تخصصي Special پژوهشي Research در دهه‌های اخیر، استفاده از مدل‌های رقومی ارتفاعی (DEM) برای شبیه‌سازی دقیق سیلاب به ابزاری حیاتی در مهندسی آب تبدیل شده است. در این میان، داده‌های ماهواره‌ای مانند آلوس (ALOS) به دلیل دسترسی آزاد و گستردگی پوشش، به‌ویژه در مناطق فاقد داده‌های زمینی، مورد توجه قرار گرفته‌اند.  با این حال، میزان دقت این داده‌ها در مقایسه با برداشت‌های دقیق‌تری همچون پهپاد در محیط‌های متفاوت، همواره محل بحث بوده است. هدف این پژوهش، ارزیابی قابلیت اعتماد به داده‌های آلوس در شبیه‌سازی هیدرولیکی سیلاب، با مقایسه عملکرد آن در دو حوضه‌ی متفاوت از نظر تغییرات سطح زمین است. دو مطالعه موردی در حوضه دروازه قرآن شیراز (با توسعه شهری و تغییرات شدید) و حوضه مارون (با شرایط طبیعی و توسعه‌نیافته انتخاب شدند. در هر دو مطالعه، مدل‌های DEM استخراج‌شده از داده‌های آلوس و پهپاد به عنوان ورودی مدل عددی HEC-RAS دوبعدی به‌کار گرفته شدند. نتایج نشان داد که در منطقه شهری، مدل آلوس به‌طور قابل توجهی در برآورد دبی اوج و زمان اوج خطا دارد، در حالی که در منطقه طبیعی، نتایج هر دو مدل به یکدیگر نزدیک بوده و هیدروگراف‌های شبیه‌سازی‌شده تقریباً منطبق بودند. یافته‌ها حاکی از آن است که داده‌های آلوس در حوضه‌هایی که طی زمان تغییرات توپوگرافی و دخالت انسانی زیادی را تجربه نکرده‌اند، قابل اتکا هستند؛ اما در مناطق شهری و دستخورده، تنها داده‌های دقیق‌تری مانند پهپاد می‌توانند بازنمایی مناسبی از شرایط واقعی ارائه دهند. این نتیجه در انتخاب منبع داده برای مدیریت سیلاب از اهمیت بالایی برخوردار است.   Introduction A key component in flood simulations is the Digital Elevation Model (DEM), which provides a numerical representation of topography and governs flow routing and floodplain delineation. In recent years, a wide range of DEM sources has become available, ranging from high-resolution Unmanned Aerial Vehicle (UAV)-derived models and LiDAR systems to ground surveys and freely accessible satellite data. UAV-based DEMs offer centimeter-scale accuracy, making them well suited for small-scale, urban, or complex terrains. However, their limited spatial coverage, operational constraints, and high costs restrict their large-scale applications. In contrast, satellite-based DEMs, such as those provided by the Advanced Land Observing Satellite (ALOS), are globally available, free of charge, and cover extensive and inaccessible areas. Despite these advantages, concerns remain about their accuracy in representing micro-topographic features, especially in urbanized basins with significant anthropogenic disturbances. This research addresses the fundamental question: to what extent can ALOS-derived DEMs be trusted for hydraulic flood modeling across basins with contrasting characteristics? To answer this, two hydrologically distinct catchments in Iran were selected: the heavily urbanized Darvaze Goran basin in Shiraz, which has experienced severe anthropogenic modifications and devastating floods, and the relatively undisturbed Maroon basin with stable topography and minimal land-use change. By comparing the performance of ALOS and UAV DEMs in two-dimensional HEC-RAS simulations, this study provides new insights into the applicability and limitations of satellite DEMs for flood risk assessment. Materials and Methods Two contrasting watersheds were analyzed to evaluate the reliability of DEMs from different sources. Darvaze Goran basin in Shiraz, characterized by rapid urban expansion, road construction, and structural interventions such as six detention dams built after the catastrophic 2019 flood, represents a highly disturbed environment. Conversely, the Maroon basin, located near Paskouhak in northwestern Shiraz, is a relatively pristine catchment of 4.3 km² with gentle slopes, loamy soils, and natural vegetation dominated by wild pistachio and rangeland species. UAV imagery was acquired using a Phantom 4 Pro drone to produce centimeter-scale DEMs. Satellite DEMs were obtained from ALOS with a spatial resolution of 12.5 meters and calibrated against UAV-derived control points. Both datasets were processed and integrated into HEC-RAS 2D. Independent grids were generated for each DEM: a 2×2 m mesh for UAV-based DEMs and a 5×5 m mesh for ALOS. Simulations were performed for multiple flood return periods in the Darvaze Goran basin and for observed rainfall-runoff events in the Maroon basin. Calibration and validation were carried out using local rainfall data, river discharge records, and Manning’s roughness optimization. Model outputs included peak discharge, time-to-peak, hydrograph shape, and flood extent. The comparison of results between UAV and ALOS provided a basis for assessing the spatial and temporal accuracy of satellite DEMs under varying geomorphological and anthropogenic conditions. Results and Discussion The results highlighted substantial differences between the two study areas. In the Darvaze Goran basin, ALOS-derived DEMs underestimated both the peak reduction and delay effects of structural flood-control measures. On average, UAV simulations indicated a 47% reduction in peak discharge and a 36-minute delay in time-to-peak following dam construction, whereas ALOS estimated only an 11% reduction and an 18-minute delay. Furthermore, ALOS simulations predicted flood routing at almost twice the travel time compared to UAV outputs, reflecting the inability of 12.5 m DEMs to capture fine-scale topographic variations. Consequently, while ALOS correctly reproduced the general trend of flow reduction, it underestimated the magnitude of structural impacts by up to 36%. In contrast, results from the Maroon basin demonstrated a strong agreement between UAV and ALOS simulations. Hydrographs generated from both datasets closely matched observed records at the local hydrometric station, with similar rising and falling limbs and minimal deviations in peak discharge. The main difference was the consistent underestimation of time-to-peak by ALOS due to lower spatial resolution, which smooths micro-topographic features such as minor channels and depressions. These findings confirm that in basins with stable landscapes and limited anthropogenic disturbance, ALOS data provide sufficiently accurate inputs for flood simulation at minimal cost. Conversely, in urbanized and topographically altered basins, reliance solely on satellite DEMs can lead to significant misrepresentation of flood dynamics. The study also revealed the importance of temporal aspects: UAVs provide up-to-date DEMs reflecting current land use and structural changes, while ALOS datasets, last updated in 2011, cannot capture recent urban growth or infrastructure development. Thus, discrepancies in the Darvaze Goran basin are not only due to spatial resolution but also to the outdated nature of ALOS data. Overall, while UAV DEMs consistently outperformed ALOS in precision, their limited scalability suggests that a hybrid approach—combining UAV data for critical hotspots with satellite DEMs for broader coverage—may provide an optimal balance between accuracy and efficiency. Conclusion This study demonstrates that UAV-derived DEMs offer superior accuracy in simulating flood hydraulics, particularly in complex or urbanized basins, by capturing detailed topographic features that influence flow paths, flood peaks, and timing. In contrast, ALOS DEMs, despite their lower spatial resolution and outdated acquisition, remain valuable for large-scale, stable basins where anthropogenic disturbance is minimal. In such contexts, they provide cost-effective and reliable inputs for flood modeling. The contrasting results between Darvaze Goran and Maroon basins emphasize the conditional reliability of ALOS: effective in relatively undisturbed terrains but prone to errors in disturbed or urban areas. The research further underscores the potential of UAV-ALOS integration as a balanced strategy to leverage the strengths of both datasets. Practical implications include improved decision-making for flood management, particularly in regions where high-resolution UAV surveys are impractical due to cost or accessibility constraints. The study also highlights the role of rainfall-driven boundary conditions in simplifying hydraulic modeling by bypassing separate hydrological simulations, and the advantages of finer temporal resolution in UAV outputs for capturing rapid flood dynamics. Future work should focus on integrating advanced UAV sensors, multi-drone coordination for larger coverage, and machine learning algorithms to enhance DEM calibration and flood prediction. By addressing the inherent limitations of both UAV and satellite DEMs, such approaches can pave the way for more robust, scalable, and cost-effective flood risk management strategies.   ماهواره آلوس, پهپاد, مدل هک-رس, سیلاب, تفکیک مکانی ALOS satellite, UAV, HEC-RAS 2D, flood modeling, DEM accuracy 91 107 http://jwmsei.ir/browse.php?a_code=A-10-1742-1&slc_lang=fa&sid=1 Masih Zolghadr مسیح ذوالقدر zolghadr.masih@jahromu.ac.ir 100319475328460018999 100319475328460018999 Yes Jahrom University دانشگاه جهرم Mehdi Sohrabi مهدی سهرابی mehdi.sohrabi1@gmail.com 100319475328460019000 0009-0006-7629-3622 No Jahrom University دانشگاه جهرم Fatemeh Rustapour فاطمه روستاپور fatemehroostapour@gmail.com 100319475328460019001 0009-0005-4334-1766 No Jahrom University دانشگاه جهرم Mohammad Reza Kargar محمدرضا کارگر Mohammadreza_kargar@modares.ac.ir 100319475328460019002 100319475328460019002 No Tarbiat Modarres University دانشگاه تربیت مدرس