Early warning for floods, landslides Carlo A. Arcilla Philippine Daily Inquirer
October 17, 2009
MANILA, Philippines—The recent “Pepeng” (international codename: Parma) and “Ondoy” (Ketsana) events and the resulting damage to life and property have highlighted problems relating to early warning systems for people to be affected by flooding and landslides.
During the Senate hearing on Oct. 14, much of the blame for the flooding in Pangasinan was aimed at managers of dams on Agno River (Ambuklao, Binga and San Roque), who apparently belatedly released large volumes of waters (5,000 cubic meters per second [m3/s]) into the river which could only accommodate 3,000 m3/s.
The glaring lack of protocols and on-site early warning systems, especially for dam operators and downstream and threatened municipalities, makes effective early warning systems a necessity.
Root cause
While much has been said about damage due to dams, less focus has been placed on the overwhelming amount of rain brought about by Tropical Storm Ondoy (600 millimeters over 6 hours) and Tropical Depression Pepeng (700 mm over one day) that was the root cause of the flooding.
It must be noted that some towns in Pangasinan were already flooded before the excess water from the dams was released, suggesting that the rainfall in the floodplain was already overwhelming the carrying capacity of rivers and dikes.
The sudden release of water from the dams only exacerbated this situation. Clearly, a rainfall-related (or technically rain-rate) forecasting system is needed to provide early warning for floods (and also landslides in high-slope areas).
Doppler: 1st-line of warning
The current system of warnings employed in the country heavily emphasizes wind velocities, with higher wind velocities meriting high signal numbers. However, this system is deficient because the predicted amount of rainfall is not included in the warnings.
Doppler radar, which is about to be installed in the Philippines, can help substantially in predicting the amount of rainfall. With this additional data, flooding models can be made much better. It will also allow the Philippine Atmospheric, Geophysical and Astronomical Services Administration (Pagasa) to issue flood signal warnings similar to typhoon signal warnings.
Tragically, Typhoon “Reming” in Bicol in 2006 caused a lot of rainfall which triggered catastrophic lahar cascading from Mt. Mayon. Reming had strong winds but the people (and the weather forecasting community) could not predict the huge rainfall that came with the typhoon. Most of the deaths were caused by lahar dislodged by the excessive rain from the volcano’s slopes.
In the case of Ondoy, there was practically no alarm for high-wind velocities (hence, most people took the storm for granted), but it packed a month’s worth of rainfall in half a day, which overwhelmed the drainage basin (or watershed) and caused flooding in downstream areas.
Knee-jerk reaction
Dams, which are human interventions to provide irrigation and electricity and help in flood control, have often been blamed for the flooding in Pangasinan and even in the Marikina Valley and in communities surrounding Laguna Lake.
This could be partly true in the Pangasinan case, but clearly it is not true in the Ondoy event which flooded the Marikina Valley and the towns and cities around Laguna Lake.
A satellite map of Laguna Lake shows that its drainage basin or watershed, including the Marikina Valley, does not have a dam within it. (See map.) Even the miniscule La Mesa Dam’s water flows away from the Marikina Valley toward Bulacan. When Ondoy struck, 600 mm of rain poured into this drainage basin within roughly half a day, which means that the whole basin, including Laguna Lake, received this rain.
The rain poured onto the slopes, cascading into the lake and causing rivers to overflow into the flood plains until the water reached the level of the lake, which had increased dramatically.
Lag time
During flooding events, there is normally a lag time between a torrential downpour and the actual rising of floodwaters in the river channels or flood plains. This lag time is critical for people living in the about-to-be-inundated areas because this is the time that they have in escaping to high ground and/or getting help.
After Ondoy, some colleagues from the National Institute of Geological Sciences or NIGS (Dr. C.P. David and his students) went around the flooded areas and tried to estimate the time when the floodwaters descended after the heavy rains. Depending on the location, they said several hours separated the time the rains peaked and the floodwaters crested.
Given these empirical data, low-lying communities can be equipped with rain gauges so they can measure rain rates (amount of rain over time).
Rain gauges
Rain gauges are devices that measure the amount of rainfall. These can be simple and inexpensive or sophisticated and relatively expensive.
At NIGS, we came up with rain gauges similar to what is used by Pagasa but we built it with locally available materials. The rain gets collected in a funnel which drains into an inner pipe where a ruler is placed.
The ruler is calibrated such that each mm of rainfall can be read directly. It is important, however, to make the users understand the concept of rain rate, which is the amount of rain divided by the time it took to reach that level.
The simple rain gauges are made of PVC pipes (although they can be improved by using transparent plastic pipes) and whose rain water collection must be observed and timed manually.
In the photos below, the leftmost part shows a complete rain gauge, which is less than one meter in height. The next photos show the funnel, inner tube and measuring stick.
In the photos above (at right), a simulation of rain shows the ruler, which is stuck on a floater above the water level, rising from the funnel, making it easy to read directly the amount of rain.
The rain gauge can be used when there is a downpour. The gauge must be put out in a place without obstruction so the rain flows freely into the gauge. The cost of materials and labor for building this simple gauge is less than P2,000.
Electronic device
More sophisticated rain gauges are electronic and data are transmitted by Wifi like the Oregon wireless rain gauge that sells for $70.
The amount of rainfall over time (rain rate) that will cause flooding and even landslides is specific to a locality, although studies by Hong and Adler (2006) in the United States have correlated rain duration and rain rate. They have come up with threshold values like 200 mm/day of rainfall rate with a potential to cause landslides (and also flooding, for low-lying areas). That rate is around 8 mm of rain per hour. (See chart.)
Who should get rain gauges?
People living in low-lying areas should get their own rain gauges so that they can calibrate the rainfall rate that causes flooding in their areas. Threatened areas like barangays in the Marikina Valley, communities around Laguna Lake and low-lying areas at the mouth of Agno River should also acquire rain gauges.
Easy to operate
The devices are easy to operate and should a heavy downpour occur over an extended period, they could provide early warning for coming floods.
Even before the Ondoy event, the Center for Environmental Concerns (CEC), an NGO, had been distributing rain gauges and training barangay folk in remote areas in General Nakar, Quezon and Rapu-Rapu Island, Albay, that are threatened with severe rains and landslides.
In June and August, CEC, together with its partner community organizations, implemented a program for developing local monitoring and warning systems for floods and landslides in selected remote villages.
The program mobilizes the organizations’ disaster-preparedness committees to operate and maintain rain gauges installed in their respective areas. They attended a two-day training, which covers the basic knowledge and skills in rainfall monitoring.
First phase
The first phase of the training is to gather rainfall data for one year, and to take note of landslide- and flood-triggering rainfall events.
The next phase is to determine rainfall conditions such as intensity, duration and cumulative amounts, which bring about floods and landslides. Data analysis will be conducted by the local organizations, with technical guidance from CEC. Further validation and refinement of critical rainfall conditions will be done as more data from the community rain gauges become available.
Recognizing the limitations of this empirical method, CEC explained to the communities the uncertainties that come with this method. Community response mechanisms will also take into account these uncertainties. While automated rain gauges and telemetry can improve the monitoring work, CEC has opted to initially use the standard-manual rain gauges due to problems of sustainability.
The villages covered by the program are located far from government centers. Oftentimes, these areas are the last to receive critical disaster-related information and support. This program bridges the large gap between local community response and the general disaster warnings from local government units and Pagasa.
Extreme weather events due to climate change are here to stay. CEC believes in the value of this program because it builds the capacity of poor communities to adapt to this environmental crisis.
It is important to note that the critical levels of rain rate, which cause actual flooding, will vary in different places. Those who obtain a rain gauge must start collecting heavy rainfall data so that they will know what critical levels of rain rate will translate into flooding in their respective areas. This site-specific warning system will bolster the general warnings issued by weather authorities.
TV, radio stations
TV and radio stations nationwide can also acquire rain gauges so they can directly read and interpret data, and warn the public about impending floods.
In Bangladesh, early warning involves mobilizing thousands of people using large speakers and megaphones. Since the Philippines is the leading texting capital of the world, we can use text messages as warning, based on locally collected impending flood data from rain gauges.
Cell sites
Early-warning preparedness would be greatly enhanced if we could connect electronic rain gauges to cell sites, which are located in mountains and have continuous power supply and data communications.
At NIGS, the Environment Monitoring Laboratory (EnviLab), which is headed by David, in partnership with several institutions, has been conducting studies and developing software and instrumentation that can be used as early warning systems for weather-related hazards.
The Automated Weather Monitoring System (AWMS) is a weather station that automatically measures rainfall, temperature and wind parameters. A similar system was designed for monitoring river and flood heights. The AWMS (and the river monitoring system) also has a communications module in which information (and warning once threshold values are exceeded) is conveyed via text messaging.
Previously designed by the EnviLab and Don Bosco (with funding from the Department of Science and Technology), the AWMS has been further improved by the UP Electrical and Electronics Engineering Department. It isnow deployed through a World Bank project in several provinces.
A design for home-based manual rain gauges is also piloted in Camarines Sur (in collaboration with Manila Observatory/Christian Aid).
In addition, the Environmental Data Manager software has been developed to aid weather satellite imagery database. It also accepts texted weather information from either manual rain gauges or AWMS.
Information and early warning save lives. In this age of the Internet, we must use all available information to forewarn us, especially the needy, of impending hazards. It is better to prevent disasters than to rescue people in distress.
Heaven sent
For typhoon and weather-related data, maybagyo.com, a website developed by an amateur meteorologist from the Bicol region has been heaven sent. People can upload rain-rate data to the Internet so that a database can be established.
In the aftermath of Ondoy and Pepeng, we must work on solutions, learn from experiences and end the blame game. With a report from Ricarido Saturay Jr.
(Dr. Carlo A. Arcilla is the director of the National Institute of Geological Sciences, UP Diliman. He holds a Ph.D. and MSc in geotechnical engineering and geosciences from the University of Illinois.
Ricarido Saturay Jr., is an MSc candidate in geology at the University of the Philippines. He now works with the Center for Environmental Concerns but was for years an instructor at NIGS.)
