O17-2020: Team H

Team H: Byebye! Landslide
Problem Definition : Understand
Problem Definition : Observe
Theory of Change : Point of View
Personas : Ideate
Prototype and Test
Final Pitch
This project (LandSwipe) aims to provide rainfall-induced landslide warning information by exploiting the use of Artificial Intelligence (AI) and involving citizen science reporters to develop digital crowd mapping. In case there is landslide warning, I wish people can have spare time to prepare themselves and keep away from landslide predicted area which lead to minimize the loss of life and materials. This project may also empowers stakeholders to systematically monitor and prepare planning for mitigation purposes.

Challenge #2: Predicting Landslide

Project Description:

Data showed an inclining trend of landslides which caused by rainfall in Indonesia from 2014-2018. The increase amount of rainfall has linear relationship with considerable number of landslide incidents. LandSwipe project aims to use Artificial Intelligence (AI) approach to predict rainfall-induced landslide pattern and involve citizen science reporter to develop digital crowd mapping. Rainfall data derived from Global Satellite Mapping of Precipitation (GSMaP), comprise of rainfall accumulation, intensity, and duration, used as input for AI models then overlay it over topographic map by using GIS. Ten configurations were performed using R, as programming language, to process the data. The performance of AI model will be tested using two statistics parameters: accuracy value and false alarm ration (FAR).

Team Introduction:

Hi I am Anistia Malinda Hidayat from Indonesia. I work as a single team and currently study about meteorology.

Wednesday, March 25 2020

Indonesian National Board for Disaster Management stated that there is an inclining trend of landslides in Indonesia from 2014-2018. Compares to the other provinces in Indonesia, Central Java has the highest number of landslide accidents. The number of landslide accidents in Banjarnegara is second highest after Wonogiri, but the population density in Banjarnegara is higher (857,12 population/km2) than Wonogiri (551,48 population/km2). Therefore landslide that happen in Banjarnegara will has higher exposure/impact to the society. Predicting landslide could become more challenging because there are so many factors that have to be considered in this project. To make it more clear, below I listed some difficulties that need to be considered in this project:

  1. In order to produce an accurate landslide prediction, I should understand what factors (that will be used as feature in AI model) that can trigger landslide in a particular area. However, there are so many factors which trigger landslides, one factor can be the most dominant factor that cause landslide in area A, but different condition might happen in area B. Several factors that can be considered as the cause are: intrinsic factors (such as type of soil, land slope, soil cover, etc) and extrinsic factors (such as human activities, eartquakes, rainfall, etc). I need to address and decide which factor that dominantly trigger landslide events in the research area.
  2. The accuracy of AI model prediction which quantified using accuracy value and Probability of Detection
  3. The availability of data that will be used as input (feature) in AI model prediction
  4. Lack of accurate data on the timing of landslide occurrences.
  5. Public awareness about the danger living around landslide-prone areas
  6. So far as I know, research on landslide prediction is such an 'exclusive' thing for researchers. Mostly, society/public doesnt know 'whats going on'. This means their research is not well-explained to the society, whereas the project was intended for them.
  7. Landslide accidents is the main problems over mountainous region and Banjarnegara is one of the complex mountainous region in Central Java province. 
  8. The availability of surface rainfall observation stations is very limited and unevenly distributed in most developing countries, especially over mountainous topography, which makes landslide prediction become more difficult to handle.
  9. Rainfall is one of meteorological elements that plays important role on triggering landslide. Rainfall data series has non-linear characteristics of data, therefore it cannot be done using classic statistical method.
  10. I do not have background knowledge about artificial intelligence, so the main challenge for further progress in the development is to develop AI model prediction and combine it with crowd information from local people
  11. The impact of extreme hydrometeorological disasters caused an increase in fatalities and significant damage to property and infrastructure around the world. This happened even though most of these disasters have been well estimated. Therefore WMO No. 1150 stated "It is no longer enough to provide a good weather forecast or warning – people are now demanding information not only about “what the weather will be, but what the weather will do”. This condition lead to the initiation concept of impact-based forecast. This project (Byebye! Landslide) is an approach to develop impact-based weather forecast technique using an artificial intelligence.

The people who potentially most affected by landslide accidents:

  1. People who live near landslide-prone area, in all ages, all gender. Landslide accident potentially causing loss of life or material. Poor people might suffer more because they dont have much money to rebuild their house, dont have health insurance when they injured, and they dont have savings for survive during that emergency period
  2. People who accidentally gather around the landslide location, whether they only pass by or doing activities near the location
  3. Government. Landslide accident might also destroy public places and facilities such as school, mosque, road infrastructure, etc which causes huge losses. Besides that, government should also prepare money for evacuation and keep their economic condition stable.



Based on the analysis on the previous explanation and my background knowledge as candidate of meteorologist, here are 3 points that I assume becoming the major cause of problems:

  1. Defining the major causes of landslide over a specific region and decide factors that will be used as feature in AI model prediction
  2. Formulate accurate AI model prediction (which model has the best performance)
  3. The way to involve community participation (citizen science reporters) and connect them with related stakeholder which can handle landslide problems

I believe that 4 points that I have mentioned before as the main cause of the problems because:

1. Some studies showed that people insist to live around landslide-prone area because:

  • Widodo et al (2017) showed that 61.6% of the people felt comfortable and at ease staying in that area. This convenience is due to environmental, economic, and social factors.
  • Laraswati (2014) said that economical factor is the main reason people still insist to live around disaster-prone areas, they can not afford more viable life
  • Head of the BNPB Information and Public Relations Data Center said that Local governments have not really regulated and enforced regional spatial planning. Spatial aspects are more effective than others 
  • Donie et al. (2007) gave another perspective related to local cultural values and religion that believe nothing can happen without God's permission.
  • Donie et al. (2007) also mentioned about the need of education about disaster mitigation

2. Aldrian and Susanto (2003) showed that based on identification of three dominant rainfall regions, rainfall pattern over Banjarnegara strongly influence by monsoonal activities. High rainfall pattern which observed during November until March (NDJFM) followed by the highest number of landslide accidents during 2014-2018. Based on Indonesian National Board for Disaster Management data, 78% (84 cases) out of 108 cases of landslide accidents happened because of rain. Therefore, I strongly believe that rainfall will be the main feature on this prediction model.

3. Automatic Weather Station (AWS) just put on Banjarnegara early in 2019 (March) and there are around 3 months data which is abnormal (the value is continuosly too high) so the accuracy of data is questionable -> using satellite observation data (GSMaP become an option)

4. News showed several scientists have made LEWS but I cannot found another news explaining us about the detail location of the instrument, how this instrument transfer/distribute the information to the society, how this instrument connect the government, related instances, and the society when disaster happen, etc. This is how the 'gap' formed, only certain or limited circles who know about this, while the users (society), who actually become target user, do not have any idea about this

Wednesday, March 25 2020

MAIN IDEAS. Since Artificial Intelligence (AI) is well-known for its capability on handling big dimension of data and process non-linear data pattern, I plan to make near-real time landslide prediction by using AI approach over a specific coordinate point area. When there is landslide warning, this information will be automatically forwarded to the society who stay within 15 km radius.

REFERENCES. This idea mainly inspires by some of these research:

  • Aleotti, P., 2004, A warning system of rainfall-induced shallow failure, Engineering Geology, Vol. 73 pp. 247–265.
  • Babich, G. and Camps, O., 1996, Weighted Parzen windows for pattern classification, IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 18 No. 5 pp. 567–574.
  • Bishop, C.M., 2006, Pattern Recognition and Machine Learning, Springer, Singapura.
  • Brunetti, M.T., Melilo, M., Peruccacci, S., Ciabatta, L., and Brocca, L., 2018, Performances of satellite rainfall products in landslide forecasting, Geophysical Research Abstracts, Vol. 20 p. 6833.
  • Ermini, L., Catani, F., and Casagli, N., 2005, Artificial neural network applied to landslide susceptibility assessment, J. Geomorphology, Vol. 66 pp. 327 – 343.
Wednesday, March 25 2020


  1. Caine (1990) and Guzzetti et al. (2008) use rainfall intensity (I) and duration (D) to predict landslide. This project add one more variable as feature, that is rainfall accumulation (A) derived from GSMAP in 2014-2018. Each of these 3 features then divide again based on its temporal scale, feature value today (D), yesterday (D-1), two days ago (D-2), and three days ago (D-3). This is due to some research that showed antecedent rainfall play important role on triggering landslide accidents (Dahal and Hasegawa, 2008). Therefore, AI model prediction (Probabilistic Neural Network) in this project will use 12 features in total to predict landslide.
  2. Besides exploit the use of AI on landslide prediction such what has been done by Ermini et al. (2005) and Saro et al. (2016), LandSwipe project also involve community participation (citizen science reporters) to give landslide prediction information and also provide near real-time landslide susceptibility maps.
  3. To make this project familiar to the user, mobile phone application named LandSwipe were developed. Indonesia is one of the 5 countries with the most social media users, it portray how society tend to be addicted to social media. LandSwipe apps was designed resemble social media but with some special features including:
  • Common people can learn and acquire knowledge related to landslide, climate change, ect -> education platform
  • Citizen science reporter can send information related to the time of landslide occurence -> used as response data in AI model prediction in this project
  • Citizen science reporter can make status/post about crack hills, heavy rainfall, or human activities that may trigger landslide and other people can give comments on this status/post -> this information will be checked and response by related intances (BNPB/BPBD)
  • Landslide victims can share their experinces, what they need, where their location, how they survive and handle things after the accidents so other people can help and learn from their experience
  • People can receive information about landslide prediction in their area and look at near real time landslide susceptibility maps around them
  • This application connect related instances, BMKG (give daily weather prediction and verify GSMaP data), BNPB, BPBD, SAR (standby in case there is warning and landslide occurences) and nearest hospitals (send ambulance to help evacuation process). People can directly send messages to these instances by using LandSwipe application
  • By assessing landslide susceptibility maps produced by this project, Local government can make further consideration such urban planning regulation, mitigation and quick evacuation regulation related to landslide, long term planning to handle climate change that potentially increase rainfall-induced landslide occurences.

There are two important things that need to be elaborated here, first is about how to develope accurate AI model prediction. A flow chart describing the way to develope AI model prediction (in this case I use Probabilistic Neural Network) is presented on the above picture. There 3 stages in general, where stage 1 consist of 2 substages.

  • Substage 1: Filtering. To make sure that the landslide accident solely triggered by rainfall, I also collect earthquakes data. If landslide accidents accompanied by earthquakes (its magnitudo range between 5.3 - 7.0 and its radii from landslide point is 40-70 km (Khazai and Sitar, 2003)) then landslide data on that day will be eliminated -> use as response data
  • Substage 2: Quality Control. Due to lack of rainfall observing data over complex mountainous region such as Banjarnegara, satellite data offers best alternative. However, before directly use the data, I try to compare its estimated rainfall value with the nearest rainfall observing station in Banjarnegara. When the data is representative, then I differentiate this value into 3 parameters consist of rainfall intensity, accumulation, and duration . These parameter then divide again based on its temporal scale (as mentioned before D, D-1, D-2, and D-3) -> use as features on AI model prediction
  • Stage 2. Pair of feature and response data in 2014-2017 used as training data, while data in 2018 used as test data. This model divide into 2 model based on the number of season in Indonesia: one model use to predict landslide during rainy season and one model use to predict landslide during dry season.
  • Stage 3. We get probability value of landslide occurences from stage 2 process. If the probability value >50% landslide is probably happen -> landslide warning issued, if <50% then landslide predicted will not occure. Besides that, we also analysis the accuracy value by compare our prediction result with landslide historical data in 2018 that we already had before (because PNN is supervised artificial neural network) and also count it False Alarm Ratio (FAR)
  • In addition to this, landslide probability value then overlay over topographic maps and it categorized into 4 labels to make it easy to understand (0-20%: NO LANDSLIDE; 21-40%: ALERT; 41-60%: CAUTION; 61-100%: LANSLIDE probabily occur) and this information will updated as often as possible (minimum daily update).

Second issue that we need to address is how to give this information to the society, how to connect all elements (society, related agencies, government, and other stakeholders), and also how to involve society as part of citizen science. The answer is to develope mobile apps (LandSwipe). The application, design, and the way it works already explained in the previous section.


LandSwipe project considered to possess an impact if it fullfill several key metrics mentioned below:

  1. Ermini et al. (2005) and Saro et al. (2016) use accuracy value to evaluate their work on predicting landslide using Artificial Neural Network. Besides using accuracy value to measure how accurate this project will be, I also plan to evaluate it using False Alarm Ratio (FAR). Some researchers said that it is important to use FAR as one of statistic parameter to evaluate prediction result. The formula to calculate accuracy and FAR shown on the above picture. 
  2. Predicted landslides are categorized as "Hit" if the predicted landslides occur followed by the real landslides occurence during observations within 24 hours from 00.00 UTC of the relevant day (D) to the next day (D + 1) at 00.00 UTC. If the observation said no landslide occur but the prediction said yes then it categorized into "false alarm". If the observation said landslide occur but the prediction said no landslide occur then it categorized into "Miss". While, if the observation said no landslide occur and the prediction said the same then it categorized into "correct negatives".
  3. If the accuracy value (degree to which pairs of forecast values and observed values correspond) is in the interval between 0 - 0.199 (very weak); 0.2 - 0.399 (weak); 0.4-0.599 (intermediate); 0.6-0.799 (strong); 0.8-1.00 (very strong) (Wilks, 2006). The higher the value of FAR, the higher the possibility of issuing wrong landslide prediction information. This comprehensive evaluation treatment will give best effort to produce accurate AI model on predicting landslide. The more accurate the model, the more the model can help reducing number of casualties, decrease economic losses, and so on.
  4. Type of ANN (AI model) that will be used in this model is Probabilistic Neural Network (PNN). The second picture (table) shows that there are 10 configurations that will be tested using the PNN model. The results of landslide prediction using PNN with 10 different scenarios then compared to determine which model that perform better in predicting landslide in Banjarnegara.

Potential people that most likely to be supportive are:

  1. Indonesian is the second most generous country in the world and there is immense potential for religious giving and private investment to support the SDGs. UN Habitat Indonesia is really potential to support this project due to the same vision related to SDGs, at least they may help me to amplify this project so the government can notice it. Because the third outcome that UNPDF 2016-2020 is Environmental Sustainability and Enhanced Resilience to Shocks and this project is very suitable for that.
  2. Yayasan Tanggul Bencana (YTB), Aksi Cepat Tanggap (ACT), and ONECARE are NGO that focus on helping victims of natural disasters. With their help, I thought I can reach government attention toward this project.
  3. I will propose this project through change.org platform because I know that Indonesians are well-known about their generousity. I will ask people to sign petition to support my project and send this petition to the government, so that they can notice my project.
  4. Researchers and collegues who have the same interest on doing landslide prediction using AI-crowdsourcing approach
  5. People who have often affected by landslides accidents
  6. People who familiar with the used of social media (as part of community participation)

Those who most likely to oppose are:

  1. People who are unfamiliar with the use of technology, including mobile phone, internet, social media, crowdsourcing
  2. People who own property over the research area and those related to developer on that area.
  3. To implement this project in society, I need to propose and communicate this project first to related government agency, local or central government. I do not know how they will response this project since it need to involve a lot of people, agency, institutions, and government so basically I can not do it alone. I need to cooperate and collaborate with them and this project will be impossible to implement if they refuse to help.
  4. Since I still become a student, I personally do not have much money or kind of 'posisition' that may ease the implementation of this project, whereas some say collaboration with related national agency, institutions, and government is complicated and I do not have any idea when that happen.


Aims: predicting landslide related to rainfall (rainfall-induced landslide)

Tools: Artificial Neural Network (PNN) involving citizen science reporters (community participation) and using of R programming language

Data: landslide historical data (response data) + rainfall data from GSMaP in 2014-2018 (feature) + citizen science reporter data (timing of landslide occurences -> validate response data)

Method: There will be two PNN model, one use to predict landslide in rainy season and the other for dry season (Indonesia has 2 season)

Target user: All people (especially people who live over landslide prone areas and familiar with the use of social media/internet/mobile phone/even AI and crowdsourcing) + connected to government/related instances/hospital

Location: One of the most landslide prone areas in Indonesia with highest population density in Central Java Province -> Banjarnegara -> if it success then will expand to another places

*Notes: detail explaination of how this project scientificly work was showed in theory of changes

What will the user experience?

Almost half of the total human population in the world are active users of social mediaThe Next Web published data per April 2018 that showed Indonesia always become the top five social media users in the world. So, I plan to design social media platform-look like that specifically created for discuss, report, comment status, or post photo about landslide information named LandSwipe. By doing this, people can enjoy social media and help other people at the same time (new way enjoying social media which is so meaningful and helpful). Additional features of LandSwipe besides its role as social media are:

  1. People can learn and acquire knowledge related to landslide, climate change, ect -> education platform
  2. Citizen science reporter can send information related to the time of landslide occurence -> used as response data in AI model prediction in this project
  3. Citizen science reporter can make status/post about crack hills, heavy rainfall, or human activities that may trigger landslide and other people can give comments on this status/post -> this information will be checked and response by related intances (BNPB/BPBD)
  4. Landslide victims can share their experinces, what they need, their location, how they survive and handle things after the accidents so other people can help and learn from their experience
  5. People can receive information about daily landslide prediction in their area and look at near real time landslide susceptibility maps
  6. This application connect related instances, BMKG (give daily weather prediction and verify GSMaP data), BNPB, BPBD, SAR (standby in case there is warning and landslide occurences) and nearest hospitals (send ambulance to help evacuation process).
  7. People can directly send messages to BMKG, BPBD, SAR, hospitals, and related NGOs
  8. By assessing landslide susceptibility maps produced by this project, Local government can take this as consideration such urban planning regulation, mitigation and quick evacuation regulation related to landslide, long term planning to handle climate change that potentially increase rainfall-induced landslide occurences.
Wednesday, March 25 2020

Who has to do what. to make it happen?

  1. The government notice, give feedback, and support the project (giving fund, mobilize people to help, ordered related agencies/instances to help, ordered expert to actualize the project)
  2. People/community offering help deliberately, without asking reward/money as return. For alternative, I will propose fulfillment of social security and health insurance for people who live over landslide area
  3. People who will be part of citizen science is cooperative and eager to learn about this crowdsourcing project or how to operate the LandSwipe apps
  4. People have the sympathy to help others through this  project and have the same interest to start making a change
  5. Scientists/researches/students will not giving up everytime their scenario is failed or facing problems
  6. Scientists/researches/students want to spend quite long time to focus on this project, even if there is no guarantee whatsoever
  7. BMKG/BNPB/BPBD sent their expert/representative to work together in this field and support this project with their data
  8. NGO help to amplify this project and communicate it to the government or news media
  9. NGO help to give sponsored or funding to this project
  10. Me, as the conceptor of this project: I will write it down into proposal/papers, propose it to my lecturers to get some advices, send it to NGOs, related agencies, and local government. I will also try to promote this on social media and post it in change.org to acquire social pressure so other people can notice it (including government and national agencies)

Who are the key partners to execute? Key partners to help others evaluate your value preposition?

  1. Collegues who have the same interest, so he/she can give feedback and help me actualize the project
  2. Experts on AI and mobile application development to help me actualize the project as my technical support
  3. Lectures who can give me insight/critics/advice about my project and introduce me to his/her network so that cooperation can be established
  4. NGOs who have more experience on how to communicate with people and also government
  5. BNPB/BPBD/BMKG and other related instances who can support us with data, give advice, and communicate with the government
  6. Last, local government who have a full control over the research area, including the permission to work on the project, the power to instruct their people, money/fund to actualize the project

What are the precipitating events?

  1. Artificial Intelligence (AI) often used by many field this recent days, applying it on meteorological to predict rainfall-induced landslide is nice idea
  2. Rigid/ well-integrated Landslide Early Warning System still not available in Indonesia
  3. The use of social media platform is very massive in Indonesia, so LandSwipe apps may change the social media behaviour (before -> only post status, now-> help other people/increase emphaty, get knowledge, direct connection to government, agencies, instances, hospital, etc)
  4. Internet is faster and well developed so it ease the actualization of this project (most of Indonesia area used 4G -> 5 G)
  5. Landslide happen more often than previous years, therefore an accurate landslide prediction is urgently need
Wednesday, March 25 2020

who else is in the field?

  1. Researchers who concern about landslide problems and do research related to it
  2. Friend and collegues who have the same interest
  3. People who live over the landslide prone areas (who potentially become citizen science reporters)
  4. BPBD (Local Agency on Disaster Management) who will be responsible if any disaster happen in the responsibility area
  5. BMKG who currently develop impact based weather forecast (rainfall-induced landslide is one of the issue)
  6. NGO who concern and working on environmental preservaton issue

What's wrong? Missing? Not working?

  1. Use additional variable as features (rainfall accumulation) compared to Caine (1980) and Guzzetti et al. (2008)
  2. Consider antecedent rainfall, many research said it plays important role on triggering landslide occurences, so I devide each three variable into 4 different temporal scales (D, D-1, D-2, D-3)
  3. Besides exploiting the use of AI (such the work of Ermini et al. (2005) and Saro et al. (2016)), I also plan to involve community participation as part of citizen scrience reporters -> to become crowdsourcing project
  4. Adapt to the latest condition, technology, and user needs/wants/habits by making personas. After succeed on building the AI model prediction (in this case I use PNN) which has the best performance on predicting landslide, I plan to develop mobile phone application named LandSwipe to distribute the produced information and provide a platform where community can have direct contact to related instances, report information related to landslide just like how they act on social media, receive knowledge related to landslide, and help this project at the same time.

Physical and intellectual resources needed (besides financial resources)

  1. Expertise on mobile device application development
  2. Expertise on Artificial Intelligence to give me insight and or critics on my method and data that will be used
  3. College friends who help and support me on communicating this issue (have good communication skills)
  4. Public/netizen to amplify this project and giving an online petition (via change.org) sign so I can propose it to the government and related agencies/instances
  5. Transportation since I am not living there and can not mobilize alone since I can not ride motorcycle properly
  6. Time to do deep research before implement it in real life

Next steps? Pilots?

  1. Learn more about the type of Artificial Neural Network that will be used in this project -> Probabilistic Neural Network. This algorithm apply Gaussian Kernel to acquire probability density function (pdf) and spread parameter (parzen window function) which values will be determined experimentally for normalized input vectors with values in the range 0-1 -> 10 configuration number/scenarios used in this project to search the which scenario model perform better.
  2. After succeed build the accurate landslide prediction AI model, I will seek for expert help to develop mobile application (LandSwipe)
  3. While building the application, I plan to write this project down into proposal or scientific papers and send this to NGO, government, and related agencies/instances.
  4. I will also promote this project on change.org to gain public attention and introduce this project to public. So, they can help/support my project by signing the petition. When I get enough public support, this can be such pressure group to urge the government or related agencies/instances to support the project too.

Cost structure? Financial Sustainability? Revenue streams?

1. Research development fund

  • Honorarium for researcher, laboratory personnel, data collectors, data analyzers, operator fees: $632
  • Stationeries, photocopies, correspondence, preparation of reports, printings, binding reports, publications, pulses, internet, laboratory materials, journal subscriptions: $1112
  • Mobile apps research and development (LandSwipe): $912 
  • Cost for travel/accomodation, sampling data/experiment cost, consumption costs: $1396
  • Maintenance costs: $1000
  • Other research support equipment: $1186

2. Marketing cost

  • Publish petition via change.org: free
  • Posting status on twitter/instagram/facebook/whatsapp to promote the project: free
  • Ask influencer to promote and amplify this project/LandSwipe apps: $698.18

*Notes: Research Budget Plan (RAB) is mostly based on the Regulation of Finance Minister Republic of Indonesia No. 78/PMK.02/2019 concerning Standard Budget for 2020

Wednesday, March 25 2020

How might this go wrong? How might the problem evolve? What are the legal, cultural and other impediments?

  1. Current AI model prediction (PNN), which has the best performance, might produce many false alarms landslide prediction due to climate change issue that may change rainfall-induce landslide pattern in the future -> solution: need regular update on choosing best configuration of PNN model (at least research has to be annually)
  2. Government or related instances/agencies do not want to help actualizing this project due to certain trust issues -> do not allow the project to be held, do not give any fund to this project, do not let their people involve in this project for free because maybe some of them still think it does not give direct economical impact/they do not receive any money when they involved
  3. When this project making a single failure over thousand times succed on predicting landslide occurences (scientific error), people (wheter society/government) can not tolerate it -> do not want to involve again on this project
Wednesday, March 25 2020

How will i promote adoption?

  1. Make petition which elaborate and tell about how this project works and ask public to sign it as form of pressure pressure to government or related instances/agencies so they can notice it
  2. Write details of this project into papers or proposal which tells the importance and urgency of this project to convince the government or related instances/agencies
  3. First option: Send proposal to social media influencer and ask their help to amplify the news related about this project, so it can reach more audiences/netizen or even the government can notice it. This recent days, becoming viral on social media is such a faster way to be notice by many people.
  4. Second option: If they (social media influencer) do not want to post it for free, then we can hire them (if we have enough money) because we can not deny that nowadays social media have a huge impact when we want to promote something (give spotlight to particular things)
  • Clarizia M, Gulla G, Sorbino G., 1996, Sui meccanismi di innesco dei soil slip. In: Int. Conf. Prevention of Hydrogeological Hazards: The Role of Scientific Research (Luino F, ed), vol. 1. Alba: L’Artistica Savigliano pub, pp. 585–597.
  • Guzzetti, F., Peruccacci, S., Rossi, M. and Stark, C.P., 2008. The rainfall intensity–duration control of shallow landslides and debris flows: an update. Landslides5(1), pp.3-17.
  • Ermini, L., Catani, F., and Casagli, N., 2005, Artificial neural network applied to landslide susceptibility assessment, J. Geomorphology, Vol. 66 pp. 327 – 343.
  • Saro, L., Woo, J.S., Kwan-Young, O., and Moung-Jin, L., 2016, The spatial prediction of landslide susceptibility applying artificial neural network and logistic regression models: A case study of Inje, Korea, Journal of geoscience, pp. 117-132.
  • Setiawan, h., Wilopo, W., Fathani, T.F., Andayani, B., and Karnawati, D. 2018. Initial development of the digital crowd mapping for landslide monitoring and early warning system, IOP Conf. Series: Earth and Environmental, Vol. 361.
Saturday, April 4 2020