Erosion risk assessment and identification of susceptibility lands

using the ICONA model and RS and GIS techniques 2 Hossein Esmaeili Gholzom, Hassan Ahmadi, Abolfazl Moeini, Baharak Motamed Vaziri 3 1 Department of Forest, Range and Watershed Management, Faculty of Natural Resources and Environment, Science and 4 Research Branch, Islamic Azad University, Tehran, Iran 5 2 Department of Reclamation of Arid and Mountainous Regions, University of Tehran, Karaj, Iran 6 Correspondence to: Abolfazl Moeini (abmoeini@yahoo.com) 7 8 Abstract. Soil erosion in Iran due to the destruction of natural resources has intensified in recent years and land use changes 9 have played a significant role in this process. On the other hand, the lack of data in most watersheds to evaluate erosion and 10 sedimentation for finding quick and timely solutions for watershed management has made the use of models inevitable. The 11 purpose of this study was to use the ICONA model and RS and GIS techniques to assess the risk of erosion and to identify 12 areas sensitive to water erosion in the kasilian watershed in northern Iran. The results of this study showed that with very high 13 slope class percentage (20% 35%) and sensitivity of shemshak formation to weathering which covers a large part of the 14 watershed, soil erodibility class is high. But there is adequate land cover along with high percentage of natural forest cover, it 15 has mitigated erosion. For this reason, the kasilian watershed is generally classified as low to moderate of erosion risk. Based 16 on the erosion risk map, results show that the moderate class had the highest percentage of erosion risk (26.26%) at the 17 watershed. On the other hand, the low erosion risk class comprises a significant portion (25.44%) of the catchment area. Also, 18 10.92% of the catchment area contains a very high erosion risk class, with most of it in rangeland and Rock outcrops second. 19 However, the erodibility of the kasilian watershed is currently controlled by appropriate land cover, but the potential 20 susceptibility to erosion is high. If land cover is redused due to inadequate land management, the risk of erosion is easily 21 increased. 22


23
Nowadays, with the growth and development of human activities, land use change, resource degradation and subsequent soil 24 erosion are major problems in watersheds. This will, in the long run, obstacle the sustainable development of the environment, 25 natural resources and agricultural lands. A study by Mohammadi et al. 2018 in Iran concluded that soil erosion in Iran has 26 increased in recent years due to the destruction of natural landscapes. Understanding the extent of soil erosion risk in the 27 absence of information in watersheds will enable critical areas of erosion to be identified. There is a lack of information in 28 most of Iran's watersheds (Naderi et al., 2011). To achieve these goals, it is useful to use empirical models using RS and GIS GIS techniques showed that the Bata area, especially in areas with high slope and low vegetation cover, there is a very serious 48 problem of water erosion. Each watershed is also important in environmental, social and economic. In this regard, by managing

57
In this study, ICONA model was used to evaluate and determine the erosion risk status in kasilian watershed in northern Iran, 58 using RS and GIS techniques to determine the impact of factors affecting erosion. The ICONA model is a qualitative one, so 59 after completing the erosion risk mapping, we performed the model validation using the modified PSIAC method. In this study,

77
For this study, OLI satellite images with 30 m terrestrial resolution and spectral bands were used. Landsat satellite images of the 78 study area were produced in July 2017. These data are automatically referenced to the UTM coordinate system and the WGS 1984 79 elliptic system during ground harvesting by known coordinate points. However, the accuracy of the geometric correction of the 80 images was evaluated by overlaying the correlation data vector on the false color images of 4-3-3 and the topographic map of 1: 81 50,000 using Gaussian filtering. The average RMSE (Root Mean Square Error) error was estimated to be 0.48 pixel geometric 82 correction, which is acceptable. The two-step process proposed by Chander et al. (2009) was used to perform radiometric correction 83 of images. Atmospheric correction is performed using the FLAASH algorithm. This program corrects atmospheric effects during 84 SWIR and VNIR wavelengths. This program uses the standard equation for spectral radiation in the sensor, which is intended for 85 solar wavelength ranges (other than the thermal range) at the Lambert levels. Rewritten images were also transcribed using the 86 nearest neighbor interpolation method.

87
Training samples were prepared to map and then supervised classification was performed. Visual interpretation methods for images 88 and maps, Google Inheritance imagery, field visits and GPS pointers have been used for this purpose. More than 20 training and 89 control samples were selected for each user class. In total, 50% of the total number of samples were considered as control points. In 90 this study, Maximum likelihood method was used, which is the most suitable method for classification with supervision and its 91 classification results are produced as user maps.

93
The ICONA model is a model developed and developed by the Spanish Institute of Natural Conservation (ICONA 1997; Bayramin 94 erosion risk, which is applicable in European countries and many Mediterranean regions and is similar to many of the effective ways 96 https://doi.org/10.5194/nhess-2020-85 Preprint. Discussion started: 20 April 2020 c Author(s) 2020. CC BY 4.0 License.
to predict erosion using RS and GIS, the model was adopted in the above countries with similar climatic conditions (ICONA 1991).

97
The ICONA model consists of seven stages, as shown in Fig. 2 this is given.  In order to prepare the slope map of the studied watershed, the digital information of maps of 1:25000 Survey Organization 1 of 110 Iran was used. After preparing the digital elevation model (DEM) of the studied watershed, the slope map was obtained in 111 ArcGIS 10-3 environment. Then the watershed slope layer is produced in five classes: low and flat slope (%0-3%), medium 112 slope (3% -12%), high slope (%12-20%), very high slope (20% -35%) And extremely high slope (more than> 35%).  The soil erodibility layer was prepared by incorporating two layers of lithofacies and slope. The erodibility map indicates the 123 1 . The use of these maps as a basic map and scientific document is free. 2. The use of these maps as a basic map and scientific document is free.

ICONA map of Erosion susceptibility
Step 3

Slope layer
Step 2

Lithological facies
Step 6 Soil protection Step 4 Land use/cover types       The results regarding different calculation steps of the ICONA model are explained as follows.

167
Step 1: Slope map 168 According to Fig. 3a and Table 2, the bulk of the study area has a very steep slope (%20-35%) of 53.8%. The high slope class 169 (12% -20%) also comprises the second tier, equivalent to 20.2% of the area. Also, the extremely high slope ranks third (17.9%).

170
While the limited surface area of the watershed is low and flat. Step 2: Lithofacies map 176 A map of the lithofacies of the kasilian watershed (Fig. 3b) shows that much of the watershed area, 72.6%, is composed of dark 177 gray and sandstone shale (Table 3). These soils are from Shemshak Formation and are of moderate to loose weathering. Also, 178 20.3% belong to Kashafrud Formation which are resistant to weathering.

179
The geological units of the area have a relatively wide range of permeability, with low permeability units having the most 180 surface expansion in the study area.

185
The soil erodibility map shows that 42.46% of the study area is highly erodible. Also, 30.58% of the watershed is in the class of 186 moderate erosivity and only 14.71% of the watershed is highly erodible. Fig. 3c and Table 4 show the erodibility status of the 187 study area.  Step 4: Land use/land cover map

200
After classifying images and producing land use maps, the classification accuracy must be specified. For this purpose, kappa 201 coefficient, overall accuracy, user accuracy and producer accuracy were calculated (Table 5). The results of Fig. 4a and Table 6 show that the highest percentage of total land use was for forest use with 67.6% followed by 205 agricultural land. Minimum land use was also in residential areas with 2.22%.

210
According to the results presented in Fig. 4b and Table 7

216
Step 6: Soil protection map 217 According to Fig. 4c and Table 8, 32.17% of the area has moderate protection. However, 30.91% of the area has very high 218 protection conditions. At the same time, only 12/11% of the area is in poor conservation conditions. Therefore, a significant 219 portion of the area is in good conservation conditions. 220 221   Fig. 5 and Table 9.     Table   239 10. According to Table 10, the highest class of erosion risk in agricultural land is in the middle class (Fig. 6a).

240
According to Fig. 6b, the highest natural forest land was in the low class (20.1%) and the highest planted forest land was in the 241 low class (3.06%). On rangeland, almost %100 of this land use is in very high erosion class. This percentage actually accounts 242 for about 5% of the total study area (Fig. 6c). Rock outcrops comprise 99% of these lands and 2.74% of the total study area of 243 the watershed is in high risk of erosion (Fig. 6d). Residential land is also located within the arable land and generally accounts 244 for less than 1% of the total study area of erosion risk classes in this land use (Fig. 6e).    weaknesses along with the impact of human factors (Fig. 7a). These factors include the type of geological formations and their 273 degrees of susceptibility to erosion, soil type, climate, surface currents, physiographic and topographic status, vegetation and

278
In this study, the validation of the ICONA model was compared with the current risk map and the current degradation map of 279 the study area with the modified PSIAC model. Finally, the erosion intensity map (Fig. 7b) is prepared and compared with the 280 https://doi.org/10.5194/nhess-2020-85 Preprint. Discussion started: 20 April 2020 c Author(s) 2020. CC BY 4.0 License. erosion risk map. In this study, the erosion and sedimentation rates were quantitatively and qualitatively determined using the 281 modified PSIAC method (Table 11). The study watershed with total scores of 53.7 and specific sediment yield of 332 tons / 282 km 2 /year with sum of scores of different factors can be said that the kasilian watershed is in the middle class in terms of erosion 283 class and in low grade in sedimentation.

303
The results show that the study area has a high slope percentage. Extreme slope class (%20-35%) with an area of 3632 ha (53.8%

304
of total study area) has increased susceptibility to erosion ( Fig. 3a and Table 2). The oldest geological units of the basin belong 305 to lithofacies of Shemshak Formation and the most recent are alluvial sediments in the rivers of the region. The Shemshak

306
Formation extends over 4800 ha in the basin. In terms of erodibility, most of the catchment area with the Shemshak Formation 307 is located in relatively erodible units and comprises more than 72% of the area (Fig. 3b and Table 3).

308
The high slope and high surface area of the Shemshak Formation, which is sensitive to weathering, has increased the sensitivity 309 and erodibility of the basin. The erodibility map of the basin indicates that a significant portion of the catchment area has high 310 erosivity (42.46%) ( Fig. 3c and Table 4). But according to Fig. 4c and Table 8

314
According to Table 6, %22.4 of the soil surface cover is represented by various human uses with the threat of human erosion.

315
The highest percentage of land use (45.9%) is in the middle erosion risk class (Table 10).

316
Also, according to Table 10, the results show that the highest and highest erosion risk classes, namely the areas most susceptible 317 to erosion caused by agricultural operations in the study area, are 645 ha in total. Therefore, with operations in the field, it can 318 be said that part of the agricultural activities in the steep slopes are unfavorable, which is very sensitive to erosion.

319
The findings showed that erosion is high in areas with high slope and low protection. This result is in agreement with the results high (rangeland) that have these special conditions (Table 10).

329
In the study area, there is a variety of erosion occurring, indicating the influence of different factors with different intensities and 330 weaknesses along with the impact of human factors (Fig. 7a). The highest erosion intensity in the region is in the low to medium  Table 11, with the assessment of erosion status at kasilian watershed, it can be concluded that the study 333 area with a total score of 53.7 and specific sediment yield of 332 ton / km 2 /year was qualitatively in moderate erosion class and 334 in terms of sediment yield. The lower class is located (Fig. 7).     Special issue statement. This article is part of the special issue "Erosion risk assessment and identification of susceptibility 367 lands using the ICONA model and RS and GIS techniques". It is not associated with a conference.