**1. Introduction**

Hydrological processes including water movement, generation of surface runoff, etc., rely upon the topography of the surface. A Digital Elevation Model (DEM) represents the bare topographic Earth surface excluding buildings, vegetation and any other objects on the surface. Some of the widely used and notable DEMs are MERIT DEM, FABDEM, etc. DEM is essential digital data for geomorphological study. Studies have been conducted for analysis of accuracy assessment of individual DEM datasets, but very limited work has been conducted wherein comparison of more than one DEM of different resolutions to analyze morphometric parameters derived from them [1]. It is, therefore, important that various openly accessible DEMs must be assessed for their suitability for morphometric analysis. Generally, it is accepted that the higher the resolution of a DEM, the more precise it is, which also indicates better accuracy and finer extraction of components of land surface and the drainage lines present [2].

A morphometric study of a watershed can be used to determine the ways in which different characteristics of an area are related. Morphometric parameter analysis is impacted by the source and resolution of DEMs [3]. Morphometric analysis of watersheds helps us to determine various aspects of linear, areal and relief parameters [4]. These parameters are used as an input for different hydrological applications, flood modelling and managemen<sup>t</sup>

**Citation:** Borgohain, A.; Khajuria, V.; Garg, V.; Koti, S.R.; Bhardwaj, A. Comparison of Geomorphological Parameters Detected Using MERIT and FABDEM Products. *Environ. Sci. Proc.* **2023**, *25*, 59. https:// doi.org/10.3390/ECWS-7-14298

Academic Editor: Athanasios Loukas

Published: 3 April 2023

**Copyright:** © 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

of watersheds. Surface water potential, areas where land degradation occurs, and flood risk can be understood through morphometric parameter analysis [5].

In the present study, we perform comparison between two openly accessible multiresolution DEMs for their efficiency in terms of geomorphometric parameters for the chosen study area.

### **2. Materials and Method**

The study area, which includes Kamrup and few neighboring districts in Assam, India is shown in Figure 1. To carry out this study, we used two openly accessible multi-resolution DEM—MERIT DEM and FABDEM, for delineation of watershed using ArcHydro tools and HecGeoHMS package in ArcMap (v 10.5). The morphometric parameters computed from both the DEMs were then computed and compared to check their efficiency for the chosen study area.

**Figure 1.** Study area map.

### *2.1. Pre-Processing of DEM*

The specifications of the datasets used in the present study are mentioned in Table 1. The two DEMs used are the MERIT DEM developed by Dr. Yamazaki, University of Tokyo by removing multiple error components from existing spaceborne DEMs (SRTM and AW3D), which represents the terrain elevations at a 3 sec resolution (~90 m at the equator), and the FABDEM, which is the first global DEM (30 m resolution) that used Machine Learning to remove building and tree height biases from Copernicus GLO 30 DEM. The tiles of MERIT DEM and FABDEM were downloaded from the following websites: http://hydro.iis.u-tokyo.ac.jp/~yamadai/MERIT\_DEM/ (accessed on 13 August 2022), https://data.bris.ac.uk/data/dataset/25wfy0f9ukoge2gs7a5mqpq2j7 (accessed on 17 Augus<sup>t</sup> 2022). They were then mosaiced and clipped according to the study area in ArcMap.


**Table 1.** Specification of datasets used.

### *2.2. Watershed Delineation*

The methodology flowchart used in this study is depicted in Figure 2.

**Figure 2.** Methodology Flowchart of watershed delineation and computation of morphometric parameters.

For the generation of watershed and sub-basins from both the DEMs, the procedure was carried out using ArcHydro tools and HecGeoHMS in ArcGIS (v 10.5).

The below mentioned steps were followed in sequential order for delineation of watershed and generation of sub-basins:




After completing the above steps, a project was set up in HecGeoHMS using the results obtained from the ArcHydro tools for further computation. As the study area includes Guwahati city and its neighboring districts in Assam, which are highly flood prone and flood occurs in these regions every year; so, out of the 9 major sub-basins extracted using the Basin merging tool, the sub-basin 6 was chosen for further analysis.

#### *2.3. Computation of Different Morphometric Parameters from Both the DEMs*

The basic morphometric parameters such as basin area, basin perimeter, etc., were calculated from the attribute table of the sub-basin. Stream order for both the DEMs were computed using Hydrology tools (Spatial Analyst tools). Basin length for the sub-basin was calculated using the measure tool and ArcHydro tool. Other parameters such as stream length, drainage density, relief ratio, ruggedness ratio, form factor, etc., were derived using standard formulas.

### **3. Results and Discussion**

Figures 3 and 4 show the delineated watershed and derived drainage network from MERIT DEM and FABDEM.

**Figure 3.** Openly accessible multi-resolution DEM. (**a**) MERIT DEM and (**b**) FABDEM.

**Figure 4.** Drainage network generated from (**a**) MERIT DEM and (**b**) FABDEM.

Different linear, areal and relief aspects including basic parameters such as area and perimeter were computed for the study area. The values of these morphometric parameters derived from both the DEMs are shown in Figure 5. It was observed that the area delineated using MERIT DEM is larger (24,628.2 km2) compared to FABDEM (21,425.6 km2). This is likely because in the study area we have chosen, FABDEM was still determined to have random artifacts and pits which were not corrected accurately [6]. Through visual interpretation, it was observed that the streamlines delineated using MERIT DEM were more accurately aligned compared to FABDEM. The sub-basin derived using MERIT DEM showed that it is governed by highest fifth-order stream, while that derived using FABDEM is governed by sixth-order stream, both indicating a dendritic drainage pattern. Although FABDEM has higher resolution compared to MERIT DEM, the total number of streams of FABDEM (493) is smaller than that of MERIT DEM (545) as the sub-basin's area delineated for it is smaller.

The bifurcation ratio values derived indicated that there is no disturbance from any geologic structures for the sub-basin. From the elongation and circularity ratio values, it was seen that the basin is oval to circular. The high basin relief values, low ruggedness number values and relatively low drainage density values obtained from both DEMs implies that less erosion occurs in the study area. It also indicates that the area has a coarse drainage texture and is associated with thick vegetation and high resistance or permeable soil. Low overland flow values (11.9 km from MERIT DEM and 13.1 km from FABDEM) computed from both DEMs also signifies that the study area is indicative of flooding conditions during heavy rainfall.

(**c**) 

**Figure 5.** (**<sup>a</sup>**–**<sup>c</sup>**) shows different morphometric parameter values derived from MERIT DEM and FABDEM.
