The Analysis of Angin Puting Beliung Risk Rate by Utilization of Remote Sensing and Geographic Information Systems in Semarang

This study examines the risk rate of Angin Puting Beliung in Semarang of Central Java. The Angin Puting Beliung is a local designation for small-scale tornadoes that occur in Indonesia, originates from differences in pressure of a weather system, leads to strong winds. Between January 2014 and December 2018, the occurrence of Angin Puting Beliung in Semarang city of Central Java reached 91 times with a total financial loss of around IDR 852,500,000 (USD 60,000). High population densities and settlements without being followed by control of spatial use and land-use change make Semarang more at risk of being hit by a tornado. This study specifically aims to determine the level of physical, social, and economic vulnerability as well as to analyze the risk level of Angin Puting Beliung in Semarang city. The survey is used as the main method in this study. Samples were taken to represent the population namely land cover, slope, and land surface air temperature with data analysis using a weighted tiered quantitative method to answer the purpose of knowing the distribution of hazard and vulnerability areas and analysis of the results of mathematical calculations to determine the risk of Angin Puting Beliung. The results obtained show that Semarang city has a high hazard level of 28,502% which is mostly found in the southwest and northeast of Semarang. High levels of vulnerability are in the sub-districts of West Semarang, Mijen, Gunung Pati, and Tembalang. The risk level with a high class ranks the least, namely from other classes with a distribution in the sub-district of Tugu and Tembalang, accounted at 16.294%.


Introduction
The data from the Indonesian Disaster Management Agency (BNPB, 2018) suggests that from 2008-2017, the Angin Puting Beliung disaster has become the second-highest occurrence after flood. This indicates that tornadoes are indeed occurring frequently in Indonesia. As illustrated in Figure 1, in 2018, there were 238 cases of Angin Puting Beliung alone in Central Java, which demonstrates the high level of exposure of the region to the disaster. Nurjani et. al (2013) states ISSN: 2807-8349 that the cyclone disasters mostly occur in the northern and southern parts of Central Java which had a relatively smoother topography than the central part, in which one area experienced a cyclone disaster, namely Semarang city, the capital of Central Java Province.  Therefore, the importance of risk analysis of the Angin Puting Beliung disaster in Semarang is due to the city's high population density, intensive use of infrastructure, and the development of industrial and business groundwork. Without being followed by spatial use control and land use change, Semarang city is at risk of the tornado disaster. In fact, the increasing population will affect changes in land use (Nurjani et al, 2013).

Sub-district
Moreover, if mitigation measures are not carried out by the government or the local community, Semarang will be more at risk of being hit by a tornado. Handoko et. al (2017) states that the most important aspect in disaster mitigation is the assessment of the vulnerability in disaster-prone areas and the method that can be used in the assessment is a combination of the Remote Sensing and the Geographic Information System (GIS) methods. The GIS is able to provide geospatial data information such as objects on the surface of the earth quickly, while providing an accurate spatial analysis system will be beneficial in analyzing disaster risk (Faizana et al, 2015). Currently, research on tornado risk is still rarely done in both national or local level, including at Semarang. One reason was due to the lack of competent resources in their fields (BPBD Semarang, 2018). We argue that studying Angin Puting Beliung disaster risk in Semarang City will facilitate the public to learn about disaster risk studies in their own zone. The area affected by the disaster, the number of people exposed to the disaster, the potential loss caused by the disaster, and the capacity it has to reduce the risk of the disaster are some of the important aspects that the public need to know. Consequently, impacts caused by disasters that potentially occur in the future can be minimized, because the community are aware and have adequate knowledge on disaster preparedness and mitigation. Based on the results of this disaster risk analysis, it is expected that the implementation of disaster management will be more effective.

Research Method
The site of this research is Semarang city, which has the coordinates of the

Data and Data Collection Methods
In this study, we use both primary and secondary data. Primary data is obtained directly by conducting data processing, direct measurement, and interviews. The secondary data is obtained from the relevant agencies. More details as follow:

Primary Data
Primary data in the form of SRTM image data obtained through the USGS website by selecting the Semarang city study area. The SRTM imagery is used to create a slope map parameter by performing image processing. Other primary data can be found in the Landsat 8 OLI Image of the Semarang study area from the USGS. Landsat 8 OLI imagery is used to create surface air temperature parameters by image processing.

Secondary Data
Secondary data in this study include spatial data as listed in

Method of Analysis
The research method used is a survey with data collection techniques carried out from population sampling. Samples are taken to represent the population. The parameters used in this study are the hazard parameters consisting of land cover, slope, rainfall, and surface air temperature. The parameters of vulnerability include physical, social, and economic vulnerability and added with the regional capacity parameters which from the results of the three parameters then analyzed to be a risk of Angin Puting Beliung.
This study will examine several variables that are the focus of research on Angin Puting Beliung disasters in Semarang. There are 4 variables that will be the focus, namely danger/vulnerability, vulnerability, capacity and risk. The variables used as input have been selected based on literature review, and taking into account the availability of data at relevant agencies.
The preparation of a hazard map is one of the important processes in determining the risk assessment of Angin Puting Beliung disaster. The hazard map contains four parameters, namely a land cover map, rainfall map, surface air temperature map and slope map. All four parameters are overlaid, then scaling and welding are carried out using the AHP method according to the influence caused by Angin Puting Beliung hazard, and combined with Angin Puting Beliung disaster data that occurs as a validation unit for Angin Puting Beliung disaster map ( Figure 1).  Determination of disaster risk mapping is performed by combining the value of hazard, vulnerability, and capacity as in Figure 4. This process is carried out using spatial calculations to produce risk maps, which is also divided into three as shown in the Figure 4.

Research Results
The results ( Figure 5)  Gayamsari which are indeed areas with settlement land cover and built land despite low rainfall in the northeast, but the area has a morphology that is flat with high temperatures. However, the area has a flat morphology with high temperatures. Whereas in the low hazard class with an area of 13,944 Km2 or 3,620%, it has relatively low rainfall and the land use is more vegetated with high density such as forests and gardens/plantations. The air temperature in the low area is not high with morphology from a bit steep to very steep.

Figure 5. Angin Puting Beliung Risk Level Map
In terms of vulnerability distribution analysis in Semarang, our findings reveal that out of the 16 sub-districts located in Semarang City, four sub-districts have high level of vulnerability.
These are West Semarang, Mijen, Gunung Pati, and Tembalang. This is because the four subdistricts have high economic vulnerability due to the small amount of productive land, especially in the sub-district of West Semarang. Later it was found that the physical and social vulnerability was classified as high because the four sub-districts have high population and vulnerable groups. This is the case especially in the sub-districts of West Semarang and Tembalang with a population of 171,000 individuals and a large number of non-permanent buildings, reaching 4,000. The four sub-districts are inversely proportional to six other subdistricts such as Tugu, East Semarang, Gayamsari, South Semarang, Candisari, and Gajah Mungkur, which get a low grade for their vulnerability. Despite the high level of social vulnerability for its density, the economic and physical vulnerability in these areas is relatively low. These six regions have a small productive land area and there are not many permanent buildings. While the rest are medium classes which consist of Ngaliyan, Banyumanik, North Semarang, Central Semarang, Genuk, and Pedurungan sub-districts. The full distribution of vulnerability levels can be seen in Figure 6. It should be noted that the level of vulnerability in an area can change to be lower or higher than before depending on the effort by the community and local government to reduce the level of disaster risk in their respective area.    (2019) Finally, the number of regions that have a low level of risk of Angin Puting Beliung in Semarang can be caused by the high level of capacity and the low level of vulnerability.
However, it is worth noting that in terms of potential danger to Angin Puting Beliung, the majority is medium. This is because a high capacity in an area allows a high level of resilience to disasters. In other words, areas with residences that already understand how to deal with the potential Angin Puting Beliung danger and participate in reducing vulnerability in the area from existing threats, can minimize the impact and losses caused by the disaster. As for the high risk of Angin Puting Beliung disaster, it can be caused by the low level of disaster capacity and the level of danger which is classified as medium to high, even though the vulnerability is classified as low to medium. This has an effect because at low capacity it can be concluded that the area has not been able to withstand the potential of a cyclone occurrence, or in this case, the population do not understand yet how to deal with and reduce the level of a tornado in Semarang. The complete spatial distribution of risk of Angin Puting Beliung in Semarang City can be seen in Figure 8 as follows.

Discussion
There have been several studies on hazard, vulnerability, capacity and risk analysis in Finally, the differences of this study with previous research are in the method and purpose. The method used is a survey method by determining the amount of accuracy of the modeling that has been done with the field conditions. This approach is more accurate in the field, using a combination of event data from the local BPBD Semarang. In its weighting using AHP, the parameters are most influential to investigate the disaster exposure. In addition, we employ the surface air temperature parameters, aimed at not only to map but also analyze the impact of the risk.

Conclusion
Our concluding notes are as follow. Firstly, the potential level of Angin Puting Beliung hazards that spread evenly in Semarang is known to almost all potential areas, with a high potential hazard known to have an area of 109,793 Km2 or 28,502% with most of it located in Ngaliyan,