Project LIGTAS: The combined forces of human and machine

by Kurt Ivan Angeles, Vanessa Martinez, and Justine Olaes

Aerial view of the massive 2018 landslide at Brgy. Tinaan, Naga City, Cebu (Photo from Philippine Star)

On September 20, 2018, a massive landslide hit Sitio Sindulan in Brgy. Tinaan, Naga City, Cebu, following days of heavy rain. It was a devastating moment as the entire area was reduced to rubble and ruins. Some residents were able to plead for help by sending text messages and calls using their phones which were buried with them. These became their lifeline as they patiently waited to be rescued out of their misery. However, many still fell victim to the disaster.

Underreporting of landslides

Most of the familiar landslides are catastrophic ones such as the 1999 Cherry Hills landslide, the 2009 Cordillera landslide, and the landslide-caused damage to the Talaingod-Bukidnon road in San Fernando, Bukidnon. These are some of the landslides that grab headlines for claiming lives and destroying properties. But to get to know more about them—and eventually learn to avoid disaster—people also need to learn about small landslides.

By observing the landslides, we can learn how they happen. Using artificial intelligence (AI), specifically predictive modelling, collecting data from small landslides and inputting them into a system can help us keep tabs on future landslides and help avoid potential tragedies.

Unfortunately, small landslides are, on many occasions, overlooked and do not make it to mass media. According to Dr. Decibel Faustino-Eslava of the University of the Philippines Los Baños (UPLB) School of Environmental Science and Management (SESAM), landslides with minimal impact are often left unreported.

“The reason siguro na hindi siya [shallow landslides] masyadong pinag-aaralan noon kasi maliit nga sila. So kapag walang namatay, ‘diba, walang [magrereport] (Perhaps the reason why shallow landslides are not being studied is that they are small. So if no one dies, no one will report it”), Dr. Eslava explained.

Project LIGTAS paves the way for Early Warning Systems

There are projects that aim to develop early warning systems (EWS) for landslides. One is the Department of Science and Technology-Philippine Institute of Volcanology and Seismology’s (DOST-PHIVOLCS) DYNASLOPE Project that collects data from media, local government units, disaster management offices, and community reports. However, the research program specifically focuses on larger landslides.

SESAM saw that creating an EWS that caters any size of landslide is a crucial step in disaster risk reduction and management. “Although maliit na landslide sila, it’s also important to know kailan sila potentially mangyari kasi they cause significant inconveniences and also sometimes, nakakapatay rin sila, nakakasira ng bahay at infrastructure (Although these are small landslides, it’s also important to know when they will potentially happen because they cause significant inconveniences and also sometimes, it can kill people and damage houses and infrastructures”), Dr. Eslava expressed.

This has led to the creation of Project LIGTAS. Also known as the “Landslide Investigations on Geohazards for Timely Advisories in the Philippines,” it is a research project spearheaded by UPLB SESAM and funded by the DOST-Philippine Council for Industry, Energy and Emerging Technology Research and Development (PCIEERD).

For nearly six years, the project has established collaborations with the UPLB College of Forestry and Natural Resources, other academic institutions, international and national agencies, and local government units in setting up equipment, conducting research and field visits, and collecting data. The project aims to measure rainfall thresholds and how they cause landslides in a particular area.

Data collection on an Automatic Weather Station (Photo from MDRRMO Mankayan Facebook)

LIGTAS’ project sites and data collection

The project started working in volcanic terrains such as Southern Tagalog and Bicol in 2018. In 2021, they expanded on highland and mineralized upland areas like Northern Samar and Benguet, and is expected to end in December 2023. Each project site has its own Automatic Weather Station (AWS) set for collecting data on minimum rainfall level that would trigger a landslide, to be monitored by UPLB SESAM’s partners in the area. One of the AWS is located at Brgy. Nagtoctoc, Lobo, Batangas, where researchers from UPLB SESAM and their partner Batangas State University-Lobo Campus conduct data collection and visual assessments of the area.

Project LIGTAS’ AWS is a rainfall gauge tool that collects real-time rainfall information in a specific area. The whole process begins once it rains. When raindrops reach the 2 to 4-millimeter (mm) minimum rainfall amount, it will be considered a rainfall event. This rainfall event will in turn be correlated with any landslide that might occur in an area. The essence is to know how many landslides have been triggered by a number of rainfall events and an amount of rainfall.

On May 26, 2023, Project LIGTAS released a landslide warning advisory at Baguio City after AWS recorded beyond the 50-mm minimum rainfall level due to Typhoon Betty. Five days later, a landslide occurred which led to the temporary closing of the Kennon Road.

An AWS has sensors that measure temperature, rainfall, relative humidity, soil moisture, and dew point (or hamog; it is the atmospheric temperature where dews form). The real-time data that these sensors have gathered will be transmitted to Project LIGTAS through the Internet or text.

Automatic Weather Station installed in Brgy. Nagtoctoc, Lobo, Batangas (Photo from Project LIGTAS Facebook)

Configuration of an Automatic Weather Station unit in Brgy. Ambassador, Tublay, Benguet (Photo from Project LIGTAS Facebook)

Currently, several AWS are located across landslide-prone areas in parts of Luzon and Visayas. In Luzon, these stations are located in Benguet (6 AWS), Lobo, Batangas (1), Catanauan, Quezon (1), and at the University of the Philippines Land Grant in Siniloan (1) and National Arts Center in Los Baños (1) in Laguna. Meanwhile, one station is situated in Visayas—in Catarman, Northern Samar.

Dr. Eslava explained that Project LIGTAS only has limited AWS installed because of barriers like cost, maintenance, and the tendency to get stolen. All AWS are purchased abroad and solar-powered, which caused people to think that the small solar panel of the equipment, the size of one to two cellphones, can power their houses.

As such, the project also makes use of weather data coming from weather stations of their partner institutions like the Project SARAi (Smarter Approaches to Reinvigorate Agriculture as an Industry), the DOST-ASTI (Advanced Science and Technology Institute), and Weather Underground.

[Flourish embed or photo slideshow] Automatic Weather Stations installed in Luzon and Visayas where Project LIGTAS collects data

Project LIGTAS as a potential forecasting system

Nonetheless, Project LIGTAS remains undeterred by the evident challenges. It is set on its track to revolutionize the country’s forecasting system so that many lives and resources would remain safe from perils. Dr. Eslava pointed out that contrary to popular belief, the country currently does not have a rainfall “forecasting” system, only “nowcasting.” Real-time updates are typically only up to six hours, therefore long-term decision-making and preparation are not possible.

The World Meteorological Organization defines nowcasting as local forecasting of the weather state from the present up to six hours ahead. Dr. Eslava explained that this is what the Philippine Atmospheric Geophysical and Astronomical Services Administration (PAGASA) does. Once the rainfall event occurs, only then will PAGASA report information on rainfall levels.

Dr. Eslava added that this is different from a study that is being conducted by their partner in Taiwan, where they are working on predicting rainfall amounts in the next two to three days. This is considered as very helpful in achieving the goals of the project, but there are other layers that need to be considered such as data validation from other partners or satellites.

Project LIGTAS aims to fill in the gaps by collecting as much data on the ground to provide input for rainfall forecasting, which can potentially help people to prepare in advance and avoid casualties. How they intend to do this involved regular citizens like you.

Where do the people come in?

This is where citizen journalism comes in, where people are provided a platform to collaborate with researchers to report information and take part in research efforts. In monitoring rainfall thresholds and potential landslides, it is not easy for the project implementers to facilitate the AWS monitoring alone.

“Hindi namin napupuntahan lahat ito. Hindi namin nakikita ‘yung mga nangyayari (We do not have the means to go to every location. We are not fully capable of personally seeing the incidents”), Dr. Eslava emphasized the importance of community’s participation.

Through the Project LIGTAS Landslide Reporting page (, everyone can help communities be prepared when landslides-even small ones-occur in their areas.

The power of the citizen’s active participation in reporting stands as a pivotal element in this joint effort. By submitting first hand accounts and/or uploading photos accompanied by details like landslide date, time, and location (coordinates to plot them on a map), the community members have the power to be the eyes and ears on the ground.

It is important that people send information on landslide events that they can provide to improve the accessibility of local weather data. This data can be used by local governments and first responders as a decision-support tool for emergency preparedness and disaster response.

This in turn allows people to contribute and understand the science of landslide studies. Through this initiative, people can learn to explore different factors such as soil condition and rainfall amount that causes landslides in their area.

However, given its user-friendly interface, the website may not be accessible to areas with weak or no internet connection. With that, Project LIGTAS also accepts reports through text messages. But while text messaging cannot be used to upload photos, the other information a text can send to the researchers is already considered helpful data.

Citizens who opt to report landslides using text messaging can send in details on a landslide event in a particular site through the project’s phone number: 0908 187 7096.

Besides community reports from the website and through text, Project LIGTAS also collects and validates data from community members and their partners, news and online reports, satellite images, and personal verifications from fieldworks.

Project LIGTAS’ types of data collected and ways to report landslides.

The collected landslide reports from these sources are transformed into a vector layer, which is then used to generate an interactive web map through ArcGIS online. ArcGIS is a web-based mapping and analysis platform that allows people to create, share, and use maps and geographic locations. Additionally, an interactive map showing the location of the reported landslide events and the location of the AWS installed by the project and other partner organizations is also present. All these features can be found on the portal, making tracking landslides easier.

Landslides come in different spectrums. No matter if they are small, big, minor, or severe, Project LIGTAS is set forth in collecting as much data as possible to provide reliable forecasting for landslides. Everyone has important roles to fulfill in making sure that hazards like landslides would be less damaging than they should be. By submitting reports to Project LIGTAS, more data will be gathered to enhance accurate forecasting and potentially improve warnings and prevention measures through a national landslide database.

Flourish embed code (Automatic Weather Stations installed in Luzon and Visayas)

<div class=”flourish-embed flourish-map” data-src=”visualisation/14178771″><script src=””></script></div>