Modeling the Relation between Area and Volume of Landslide(Case study: Neyshabour Baqi Basin)

Introduction In Iran, landslides cause many problems every year, such as: destruction of roads, destruction of agricultural lands and pastures, residential areas, soil erosion and transition a great deal of deposits to the watersheds. It is important to be aware of number, area and volume of landslides in order to estimate sensitivity, determination of landslide danger and long-term evaluations to estimate bulk sensitivity. Area and number of landslides information is simply achieved through aerial photos, satellite images and field inspections. However, these methods are of no use for volume determination. It is a difficult task that needs surface and subsurface geometrical data from the rapture slope. Gathering this information which is performed as field operations is difficult and expensive and estimation of the volume of slope in steep hills is very challenging. Thus, estimation of landslide volume may be only carried out by taking experimental relations that connects the volume to geometrical raptures measurements especially the area.Since, deposit delivery in a basin exit is very important in watersheds management and most of deposits are caused by landslides on side-lines of rivers; the importance of estimation of landslides volumes has to be emphasised. Hence, the present study tried to find an experimental relation to estimate landslide volume in Neyshabour Baqi Basin (southern slopes of Binaloud) considering the significance of deposit volume caused by bulk erosion and also expensive and time-consuming field operations. It expands from 59” 38 ’58° to 13” 44’ 58° longitudes and from 09” 31’ 58° to 30” 38’ 36° latitudes; in southern slops of Binaloud Mountains. 3. Materials and methods In order to suggest an experimental relation to calculate the volume of recorded landslides, the only ones were chosen that had complete area and volume information. Thus, data of 44 observed landslides including longitude, latitude and area was given to SPSS Software to set a relationship between area (AL) and volume (VL). In order to model the experimental relation of AL and VL, considering existing statistical relations, general form of this equation was used: Finally, in order to estimate the volume of landslide by area information considering regression assumptions, a power regression model was obtained. In this relation, the observed volume is given by multiplication of slide area by observed depth obtained from remote calculations and GIS. Various experimental power relations which were used for landslide volume calculations, number of landslides and the associated area with calculated volume are given in table 1. Relations were applied for 44 observed landslides areas in the basin and results were compared with the observed values. At last, a statistical improved model for Baqi Basin was also compared with other relation. Comparison was made through R2 coefficient, percentiles statistics values, maximum, minimum and average and the root mean square errors (RMSE). 4. Results After collecting the data required from surveys, first their accuracy and quality were controlled and then their statistical properties were calculated and provided. According to the results of this table, landslides in the Baghi area are in a relatively wide range of area, size and depth, such that their area (AL) was in the range between. The observational volume or size of landslides (VL) was in the range between and their depth (DL) was in the range of. This wide range may increase the standard deviation (SD) followed by variation coefficient for each parameter. Because there are no observational data for depth, we decided to calculate the observational depth of landslides by remote sensing and GIS. In addition, to ensure the reliability of data, by selecting 8 sites for landslide out of 44 landslides occurred in the region and field studies, the accuracy of calculation depth were estimated indicates No. 44 of landslide with complete data of area (AL) and volume (VL), in a diagram with Log-Log coordinates. the frequency of slides is higher in the area range between 100000 m2 to 1000000 m2 and volumetric range between 500000 m2 to 5000000 m3.. Exponential regression fitness on the data of observational area and volume finally resulted in creation of equation (1). To evaluate the calculation values of size of landslide by related equation, these values were compared to volumetric estimations by other equations. After calculating the size of landslides in the Baghi area, the diagram for range of area determined were drawn by different equations. it can be said that equations like Imaizumi and Sidle (2007), Imaizumi (2008), Rice and Foggin (2008), Rice (1969) were relatively in good conformity to diagram of this study (equation 1); however, equations developed by Larsen and Sanchez (1998), Guthrie and Evans (2004), Marrtin et al (2002) have relatively lower estimation than developed equation. 5. Conclusion Results from this study indicated that the equation developed for Baghi area are in good conformity to some equations developed by others internationally. Because these equations presented in study areas with different local and physiographical conditions than Baghi are as well as for different ranges of areas, therefore, this conformity indicates that the relation between volume and area of landslide is basically geometrical and independent from local and physiographical conditions. In final conclusion, according to statistical accuracy of model provided and its conformity to some other equations as well as benefiting from results of this model from data obtained through whole area, this model can be considered as a desirable model for Baghi area and for calculating the size of landslides in the Eastern North of Iran.