نوع مقاله : مقاله پژوهشی
نویسندگان
1 دپارتمان جغرافیای طبیعی، دانشگاه اتوش لورند، بوداپست، مجارستان
2 دپارتمان جغرافیای طبیعی، دانشگاه اتوُش لورند، بوداپست، مجارستان
چکیده
کلیدواژهها
عنوان مقاله [English]
نویسندگان [English]
Introduction: Digital Elevation Models, also known as DEM, are widely utilized tools in Earth sciences, particularly in geomorphological studies. Although various DEMs with distinct characteristics exist, only a limited number are freely accessible to researchers. Furthermore, not all DEMs exhibit identical performance. Factors such as pixel or cell size and the algorithm employed in DEM generation are among the most critical elements influencing their performance. Therefore, one of the most crucial considerations in employing DEMs is selecting the most suitable and optimal model for geomorphological studies. In this context, the present study aims to identify the most appropriate DEM by evaluating four commonly used and freely available models. Ultimately, based on the performance of each model, the most suitable DEM for geomorphological investigations will be determined. Selecting an appropriate DEM will contribute to producing high-quality studies and obtaining reliable results in geomorphological research.
Materials and Methods: This study utilized four freely available DEMs including SRTM, ASTER, AW3D30, and ALOS PALSAR. The spatial resolution of the first three models is approximately 30 meters, while that of ALOS PALSAR is 12.5 meters. To quantitatively assess the performance of these DEMs, the study employed parameters such as the Root Mean Square Error, geomorphons, slope, hypsometric curve, and cross-sectional profile. By selecting these parameters, the DEMs were evaluated from various aspects, including vertical accuracy, horizontal accuracy, and the impact of cell size. Additionally, to mitigate errors inherent in the DEMs, the Fill operation was applied to all models used in the study. The fill sink tool is available in most of the commonly used GIS software.
Results and Discussion: According to the RMSE index, the AW3D30 model exhibited the lowest error with a value of 6.77, followed by the SRTM and ALOS PALSAR models with values of 7.36 and 8.02, respectively, while the ASTER model showed the highest error with a value of 11.95. The comparison of geomorphon classes revealed that the ALOS PALSAR model produced a total class area of 120.76 km², which is the closest estimation to the actual area of the region (121.5 km²). In contrast, the other three models yielded similar areas of approximately 119.9 km². The results further indicated that the SRTM model provided landform classifications, including valleys with an area of 10.6 km², that were closest to those derived from the ALOS PALSAR model (8.96 km²). The analysis of slope values also demonstrated that the ALOS PALSAR model recorded the steepest maximum slope, equivalent to 74.73°, whereas the maximum slope values in the SRTM, AW3D30, and ASTER models were 68.99°, 68.37°, and 66.36°, respectively. The hypsometric curves likewise showed that the ALOS PALSAR and SRTM models exhibited similar performance; for example, at a relative elevation of 39, the corresponding relative areas were 94.80 and 95.94, while the corresponding values for the ASTER and AW3D30 models were 88.84 and 89.47, respectively. The cross-sectional profiles indicated that the ASTER model had the lowest vertical accuracy and tended to overestimate elevation values compared to the other models.
Conclusion: In this study, a set of tools was employed to assess and identify the most suitable DEM for geomorphological studies. Based on the findings, despite the superior spatial resolution of the ALOS PALSAR model, the SRTM model emerged as the most appropriate freely available DEM for geomorphological applications, particularly in Iran. The primary reasons for selecting this model include its nationwide spatial coverage and its performance, which closely approximates that of the ALOS PALSAR model. Additionally, one of the key limitations identified in this study was the absence of high-precision local models for result validation. Consequently, for future studies, it is recommended to utilize a high-resolution DEM to enhance the accuracy and reliability of validation processes.
کلیدواژهها [English]
)