The Identification of Glacial Cirques of ZardKuh Based on Morphometric Features

Abstract
Introduction
Glacial cirques are one of the most important erosional forms in highlands. Landforms are the basis of geomorphology studies. Therefore, depending on the purpose of different researchers, it has been defined differently. It provides an evidence of effective processes and features on the earth surface during the past and present era (Etzelmüller, B., Sulebak, J.S., 2000). The basic principle emphasized by the geomorphometry is the existence of a relationship between the shape and its related numerical parameters for describing landforms. Numerical geomorphology studies spatial and statistical features along with the relationships and patterns of the points (Evans, 1972). The automatic classification of geomorphologic units and landforms is mainly based on the morphological parameters (Giles; Franklin, 1998). The morphometrical parameters represent the shape of the earth and the procedures creating them (Jamieson, 2004). On the other hand, the basis for classifying the units in geomorphology is based on the hierarchical theory (Ramesht, 2006). Different methods exist for providing digital elevation model and simulation of surface effects which can be used in different geosciences such as earth hazards, erosion, geomorphology, ecology, hydrology, and other related fields. The efficiency of the geomorphologic indices in dry areas was evaluated by the Wood method (Shayan et al., 2012). Makram et al. (2014) used a topographic index at any point of the digital elevation model to extract landforms. The purpose of this research was to provide a semi-automatic method for the detection and extraction of glacial cirques in Zardkuh Bakhtiari area based on Wood and Evans methods.
Methodology 
In order to achieve the research objectives, a 20m digital elevation model was generated from the 1: 25000 topographic map. Then the first (slope) and the secondary derivatives, a plan and profile curvatures, general curvature, the minimum and maximum curvatures, Longitudinal and Cross-Sectional Curvatures were extracted in ENVI software from the first derivative (slope). The data from the first and secondary derivative layers were standardized as per fuzzy logic, resulting in a single RGB map.  By combining the bands of the map, it’s possible to produce significant color outputs. Accordingly, the cirque-like shape holes were identified and extracted manually from the topographic maps. In order to conduct a controlled classification, four cirques were introduced in the RGB map as sample cirques and then the semi-automated model was implemented in the GIS software to find other cirques which were similar to the sample cirques. The Evans and Cox (Evans, 1974) proposed model was used to extract the cirques using the focal point command and the model output was adjusted by supervised classification. Finally, the general accuracy of the classification was evaluated by using a cross-validation method.
Result
In order to extract the glacier cirques in the Zardkuh area, 26 cirque like holes were manually identified on topographic and slopes maps. Since the purpose of this study was the extraction of the glacial cirques based on the morphometric parameters and characteristics, seven morphometric indices including profile and plan curvatures, minimal and maximal curvatures, longitudinal and cross-sectional curvatures, and general curvature of the region derived from the first derivative or the slope map of the region were used. Then, eight morphometric layers were standardized and combined as per fuzzy. Through layer stack, a RGB map containing all morphometric parameters was created. By changing the bandwidth of this map, significant color outputs, such as the display of aretes, talwegs, slop aspect, height differences, and the like can be obtained. To run the supervised classification model, the morphometric characteristics of four developed cirques were extracted. The values of each of the geomorphometric parameters represent the characteristic features of the landforms. Subsequently, four developed cirques were introduced as a training circus on the RGB map derived from the combination of the morphometric parameters. The results of the supervised classification represents 14 cirque like holes out of 26 holes. Here, due to the complexity of the geology and the high-precipitation region, all holes specified by the MLC model were examined using the cirques classification system introduced by Evans and Cox (1974). From the adjustment of the MLC model and the Evans and Cox definition, by observing the cirques and examining theories of experts, it can be concluded that there are eight developed cirques in the studying area.
Discussion and Conclusion
The purpose of this research was to provide a semi-automatic method for the detection and classification of glacial cirques landforms in Zardkuh Bakhtiari area. In this study, for each cirque, geomorphometric indices including plan, profile, general, minimum, maximum, cross-section, and longitudinal curvatures were extracted and calculated. The results of this study showed that Zardkuh glacial cirques do not have the most common shape of cirques due to being formed on carbonated formations, but the geomorphologic indices have greatly shown the quantitative and qualitative features of the cirques in the Zardkuh area. By using the MLC model approaches and by conforming its output with the definition of the cirques presented by Evans and Cox (1974), eight fully developed cirques in the studying area were identified.
The main reason for this is the carbonate lithology structure of Mount Zardkuh, which has removed the typical shape of the cirques due to the dissolution. Therefore, it can be suggested that the geomorphometric approach in identifying the automation of circuses in these areas cannot be very effective, but it can have a lot efficiency in visual interpretation and identification of landforms.