Landslide disaster mitigation and adaptation strategy in one of the East Java horseshoe areas using geographic information system analysis

The horseshoe area of East Java is an area that has abundant natural resource potential. In the northern part of the horseshoe region there is a very good biological natural resource with the potential to be used as a food source. Meanwhile, in the southern part is rich with biological and mining potential. This potential causes changes in land use in the northern and southern regions which can trigger landslides in the Horseshoe area. Changes in land use that change the topography and vegetation to become steep and barren will have a very significant impact on landslides. This study aims to map and predict areas affected by erosions using environmental and physical data using Geographic Information System (GIS) Analysis. The results of the mapping are areas that have threats and are vulnerable as well as capacity to erosion. Based on the results of the mapping, it shows that the index of threats greatly influences the hazard of landslides that occur in the area. Based on the result of the mapping, it shows that the disaster mitigation and adaptation strategy is to reduce the threat index or very high vulnerabilities that affect the risk of landslides that occur in an area.


Introduction
Horseshoe is the name of an area in the eastern part of East Java Province.This area covers seven districts (Banyuwangi, Bondowoso, Jember, Lumajang, Pasurun, Situbondo and Probolinggo).The Tumpangpitu area is part of the horseshoe located in the Pesanggaran sub-district, Banyuwangi.This area has a variety of potential natural resources that can be exploited, especially the potential of mineral resources.Observation of the geological structure using lineament analysis of satellite data has the potential to control the presence of these mineral resources.Furthermore, the effect of the existence of lineaments can be tested on the structure geology and landslides.
The geological conditions in the Tumpangpitu area make it rich in springs.However, this causes the Tumpangpitu area to become prone to landslides if the rock experiences cracks which results in rock wear on soft lithology and is reactive to erosion processes.Therefore, it is necessary to map landslide-prone zones so as not to cause casualties.
Structural geology is an important factor for regional development because it acts as a control system for the presence of potential mineral resources, hydrogeology and weak zones [1].The existence of this pattern also affects the hydrogeology when the fault structure intersects the rock layers [2].In addition, the geological structure of the joints affects the strength of the surrounding rock, so it is important to pay attention to the relationship between the joint pattern and the surrounding alignment [3].Several studies on landslides have been carried out using vegetation density [4][5][6][7] but the relationship between vegetation density and lineament that occurs in an area has not been connected to landslides.Even though alignment is a trigger for landslides in an area.In this study, satellite imagery data and other parameters were used to process the alignment pattern and then the data was validated in the field.The purpose of this study was to obtain lineaments and vegetation patterns as well as geological structures in determining landslide disaster mitigation and adaptation strategies using satellite imagery and SRTM DEM data in the Tumpangpitu Banyuwangi area as part of the Horseshoe area.

Methods
Identification of disaster threats is an effective mechanism for providing an overview of disaster risks in an area by analyzing geological conditions that can cause potential hazards and their impacts, such as exposed populations, loss of property and livelihoods, and damage to the environment on which they depend.Identification of disaster threats is carried out to identify hazards from an area which is then analyzed and the possibility of risks arising.In addition, the identification of disaster threats also analyzes the mitigation and adaptation that must be carried out by the government and all levels of society in disaster management efforts, as well as determines recommendations for disaster management policy directions to address disaster reduction that has been identified.
The research uses two approaches, namely remote sensing and geostatistical approaches using geographic information system-based analysis.Remote sensing approaches are used to obtain information about the level of vegetation density through analysis of data obtained using tools without direct contact with the objects, areas or phenomena being observed [4-5][8].How remote sensing works is shown in Figure 1.Furthermore, to minimize the occurrence of errors in spatial data processing, Ordinary Kriging is carried out.This geostatistical interpolation method is often associated with the acronym BLUE (Best Linear Unbiased Estimator) in order to achieve optimal estimation.The Ordinary Kriging method is carried out by considering the spatial variation of the data as a function of the variogram or covariance.This method relies on the assumption that variation is random and spatially independent, and the random process is intrinsically stationary with a constant mean and variance that only depend on the distance between data [9].The kriging system equation and its variance are as follows: The OLI imager sensor (Operational Land Imager) on the Landsat-8 has 1 band infrared near and 7 band looked reflective, will cover wavelength electromagnetic radiation reflected by objects on the surface of the earth, with resolution spatial 30 meters.The OLI imaging sensor is capable of both spatial resolution and spectral resolution which resembles the ETM+ (Enhanced Thermal Mapper plus) sensor from Landsat 7 or 8 imagery.However, OLI imaging sensors do not have a thermal band.However, this OLI imaging sensor has new bands.Those are band-1: 443 nm for detection aerosols line beach and band 9: 1375 nm for detection cirrus [10].
The lineament is a linear surface feature that can be mapped and differentiated from the pattern of surrounding features.Straightness can be (1) straight river and valley segments; (2) parallel surface indentations and depressions; (3) changes in soil color that reveal variations in soil moisture; (4) stripes on vegetation; (5) changes in the type and height of vegetation; (6) sudden topographic changes.All of these phenomena may be the result of structural phenomena such as faults, series of joints, folds, cracks, or faults [7].Satellite images will be better at distinguishing lineaments than aerial photographs because they use various wavelength intervals of the electromagnetic spectrum [11][12].Automatic lineament extraction methods, such as the Segment Tracing Algorithm (STA), identify straight paths in satellite images by examining local variations in the degree of gray [13].
The lineament density is calculated by dividing the study area into grids, calculating the lineament parameters for each grid, and making lineament density contours [14].This lineament density map helps identify areas with intensive tectonic activity, visualizes the density of faults and fractures with good permeability, and shows zones of structural openings (jogs) as potential pathways for emerging fluids so that they can become weak or sliding zones [15].An illustration of the lineament density is shown in Figure 2, where the center of the grid or dotted line is the center of density while the thick line represents the density outside the grid and the thin line represents the alignment within the grid.Image satellite which used were Landsat-8 and DEM SRTM.SRTM DEM is used to determine the lineament of morphology and geological structure.The remote sensing process involves two stages, namely data acquisition and data analysis (Figure 1).The data acquisition phase involves energy sources, energy propagation through the atmosphere, energy interactions with objects on the earth's surface, and energy returns to the sensor.Sensor data is then generated in the form of images or digital.Satellites are equipped with sensors that can inform the types of minerals present [4].
Research location is in Region overlapping, Pesangaran, Banyuwangi which is marked with a square as shown by Figure 3. Geographically, this area is located at 8° 35' 20'' S and 114° 01' 08'' N.This research requires the preparation of tools and data which are divided into two groups: tools for research and data for data analysis.The research tool is the QGIS software.Software is used to process hazard map data taken from BNPB (National Agency of Disaster Management) .Other data employed in this research is comprised of satellite imagery data (Landsat-8) and DEM SRTM.Image processing of Landsat-8 satellite data was carried out using Qgis which was classified to determine the density of vegetation, while DEM SRTM processing was carried out using Matlab 2015 to analyse the straightness of the morphology and geological structure in the research area.Landsat 8 satellite data is adopted to map vegetation density using the NDVI formula below: The data processing includes the extraction of straightness and vegetation density.The DEM data was employed to identify lineaments in the research area.The DEM data is converted into multishaded relief and the lineaments are calculated using the Segment Tracing Algorithm (STA).The density of lineament is determined by considering the spatial correlation based on geostatistical methods.Landsat-8 satellite imagery corrected for the atmosphere to determine the density of vegetation using image classification based on band values.The results are interpreted taking into account the lineament and landslide data from BNPB.

Results and Discussions
The results of liniament processing show that the straightness density varies greatly in direction.Based on the alignment map in the overlapping area, it shows that there are narrow and elongated alignments in the northeast and southwest directions as shown in Figure 4. Liniament data processing comprises lineament calculation automatically applying the Modified Segment Tracing Algorithm (MSTA).Before processing, the DEM data was converted into a multishaded relief, then followed by the computation of straightness applying MSTA.Based on the analysis as illustrated in Figure 4, the lineament in the study area is very tight.Apart from the northeast and southwest directions as indicated yellow lines from the interpretation, the lineaments in the overlapping Tumpangpitu area are also dominated by the northwest-southeast direction as showed red lines.This shows that the sliding plane is a lineament direction and the forces that form the geological structure come from the north or south.To obtain a landslide map as shown in Figure 5, data from BNPB in the form of an arc map is extracted using the mask feature and reclassified using ArcGis and then processed using QGIS.Before being processed in Qgis, raster data must be converted to polygons using ArcGis.From the classification results, it was found that the landslide zone that occurred in the overlapping area showed a high level of threat.
To reduce the risk of landslides, areas with zones with a high level of threat must be avoided, especially in the western and southern parts of the Tumpangpitu area.
In addition to lineament data to determine slide adaptation in the Tumpangpitu area, it is necessary to know the relationship between vegetation density and sliding.The Vegetation Density Map was obtained from the classification using channel 5 of Landsat satellite imagery and channel 4 of Landsat 8 satellite imagery in 2022.The results of the display of the landslide hazard and vegetation density maps show that there are zones in the Tumpangpitu area that have medium-high vegetation that can be occupied in the Tumpangpitu area.Based on the map in Figure 6, it can also be seen that most of the landslide zones are areas with medium-high vegetation density.This means that areas with medium-high vegetation experience the potential for landslides due to land changes.This also means that the area has severe geology structure that can trigger landslide in that area.

Conclusion
From the results of research on the lineament and density of vegetation against the threat of landslides in the Tumpangpitu area, it can be concluded that the geology structure pattern in the overlapping area has a west-northeast-southwest direction, and a northwest-southeast direction which is the slip plane.The slides plane originating from geological structures require mitigation by reducing activities in highlevel threat zones.The adaptation strategy is carried out by relocating near the area but to a low-level threat zone.

Figure 2 .
Figure 2. The relationship between lineaments and grid boundaries as well as the center point of the grid.

Figure 6 .
Figure 6.Landslide disaster hazard map and vegetation density.