A.M.F. Lagmay1,2, J. Mendoza2, K. Punay2, N.E. Tingin2, P.A. Delmendo2, F. Cipriano2, J. Serrano2, L. Santos2, G. Sabio2, and M.A. Moises2
1National Institute of Geological Sciences, University of the Philippines, C.P. Garcia corner Velasquez street,
U.P. Diliman, Quezon City. 1101 Philippines
2Nationwide Operation Assessment of Hazards DREAM flood modeling component, U.P. NIGS, C.P. Garcia
Ave., U.P. Diliman, Quezon City, 1101 Philippines
Correspondence to: A.M.F. Lagmay
[email protected]; [email protected]
Urban floods from thunderstorms cause severe problems in Metro Manila due to road traffic. Using Light Detection and Ranging (LiDAR)-derived topography, flood simulations, and anecdotal reports, the root of surface flood problems in Metro Manila is identified. Majority of flood-prone areas are along the intersection of creeks and streets located in topographic lows. When creeks overflow or when rapidly accumulated street flood does not drain fast enough to the nearest stream channel, the intersecting road also gets flooded. Possible solutions include the elevation of roads or construction of well-designed drainage structures leading to the creeks. Proposed solutions to the flood problem of Metro Manila may avoid paralyzing traffic problems due to short-lived rain events, which according to the Japan International Cooperation Agency (JICA) costs the Philippine economy P2.4 billion pesos/day.
Keywords: street flood, urban flood, LiDaR, Metro Manila, flood, traffic
Metro Manila is located on an isthmus between the Manila Bay, which opens to the South China Sea, and Laguna de Bay, an approximately 900-square-kilometer, 2-meter-deep freshwater lake. The entire region is composed of one major catchment called the Marikina River Basin, which covers 535 square kilometers, and eight smaller, river sub-basins, which cover 683 square kilometers that drain directly into Manila Bay and Laguna de Bay. The Marikina, Pasig, San Juan and Tullahan rivers serve as the main outlets for a network of tributaries of the Marikina River Basin and smaller catchments of Metro Manila (Figure 1). Highly urbanized and populated by almost 12 million residents (Cox, 2011), the metropolis lies on one of the widest floodplains in the Philippines.
In 1986, the Manggahan floodway was built to reduce flooding along the banks of the Pasig River by diverting floodwaters into Laguna de Bay. A complementary project called the Napindan Hydraulic Control System was built at the confluence of the Marikina and Pasig rivers to regulate tidal flow and prevent intrusion of polluted water into Laguna de Bay. Despite these flood control measures, extreme floods are still common in Metro Manila, generally occurring at least once a year during the rainy season. Residents have known about the region’s flood risk since at least the Spanish colonial period (1521 to 1898). But usually when floods strike Metro Manila, they are small, affecting only a few neighbourhoods at a time (Zoleta-Nantes, 2000).
Apart from devastating floods like those spawned by Tropical Storm Ondoy in 2009 (Lagmay et al., 2010) and the typhoon-enhanced southwest monsoon rains in 2012, 2013 (Lagmay et al., 2014) and 2014, which brought the entire metropolis to a standstill, more frequent and short-lived thunderstorms that affect portions of Metro Manila also paralyze the nation’s capital because of the heavy traffic they cause. Once parts of the road network of the metropolis are blocked due to half-tire deep floods or higher, mammoth traffic develops and paralyzes the entire city. According to a JICA report, traffic jams due to thunderstorm-related flashfloods cost the Philippine economy PhP 2.4 billion a day from waste of gasoline during traffic and lost opportunities (Rodis, 2014).
Flashfloods are traditionally blamed on aggravating factors such as the loss of infiltration capacity due to urban concrete, a century-old drainage system and streams clogged by garbage. These factors are real and indeed contribute to the worsening flood problem of Metro Manila. This study presents a reanalysis of the nuisance flood events caused by brief but heavy downpour in the metropolis to identify other and possibly bigger culprits with the aim to find a relatively inexpensive solution to the costly traffic nightmare.
The Metro Manila Development Authority released its initial list of flood-prone areas in the nation’s capital (Table 1). This list was plotted on a map by netizens and replotted in this study by overlaying the flood-prone areas with the flood map produced by the Department of Science and Technology (DOST)-Project NOAH. Twitter accounts of floods from thunderstorm and southwest monsoon rainfall events were documented to check consistency of the MMDA reports with actual flood events documented in social media.
|1. España – Antipolo – Maceda
2. P. Burgos (City Hall)
3. R. Papa, Rizal Avenue
4. Buendia Extension – Macapagal Avenue
5. Buendia – South Superhighway (northbound)
6. Buendia – South Superhighway (southbound)
7. Osmeña Skyway (northbound)
8. Osmeña Skyway (southbound)
9. Don Bosco
10. EDSA Pasong Tamo, Magallanes
11. West Service Road, Merville
12. East Service Road – Sales Street
13. McKinley Road
14. C-5 Bayani Road
15. C-5 BCDA
16. C-5 Bagong Ilog
17. EDSA – SM Megamall
18. EDSA – Camp Aguinaldo Gate 3
19. Quezon Ave. – Victory Ave. / Biak na Bato
20. NLEX – Balintawak Cloverleaf
21. North Avenue fronting Trinoma Mall
22. EDSA – North Avenue
23. Philcoa Area
Table 1: MMDA list of flood-prone places in Metro Manila
After identifying the 23 flood prone areas, analysis of LiDAR-derived digital surface models (DSMs) and digital terrain models (DTMs) from the CSCAND project was conducted to compare the profiles of the roads and road side. Two additional roads that intersected with creeks were examined in the University of the Philippines campus in Diliman, Quezon City. These are A. Roces Street and C.P. Garcia Avenue near Katipunan Road. A. Roces Street is in the center of the campus, built as part of the development plan of the UP Diliman in the early 1900s while C.P. Garcia was built in the late 1970s and early 1980s. Field work was also conducted to check the drainage crossing the streets in those areas.
Flood simulations over LiDAR topography with 1 x 1 m pixel resolution and 0.15 m vertical accuracy in the sub-basin area of interest was then conducted using Flo-2D GDS PRO, a GIS integrated software tool that creates an integrated river and floodplain model by simulating the flow of the water over a system of square grid elements. The software utilizes the St. Venant equations for continuity and momentum equations (Equations 1 and 2) in computing the velocity of the water across the grid element boundaries.
where V is the average velocity between two grids, h is the flow depth, and i is the excess rainfall intensity which may be nonzero. Equation 1 shows the continuity equation wherein V is a dependent variable while Equation 2 shows the full dynamic wave form of the momentum equation. The variables in the momentum equation include the friction slope (Sf), bed slope (So), acceleration due to gravity (g), pressure gradient , and the local , and convective accelerations. These differential equations are solved using the finite difference scheme to get the velocity across the boundaries in the eight potential flow directions (north, east, south, west, northeast, northwest, southeast, and southwest) of every grid in the model.
The floodplains were delineated into sub-basins based on elevation values given by the DTM. Each sub-basin is marked by ridges dividing catchment areas. These catchments were first delineated using a set of tools compiled into a single processing model. These tools allow ArcMap to compute for the flow direction and accumulation of water over the catchments based on the elevations provided by the DTM. Once processed to yield flow direction and accumulation, delineated floodplains were further refined. The entire area was subdivided into sub-basins in such a way that it can be processed properly. This was done by grouping the catchments together, taking special account of the inflows and outflows of water across the entire area.
After loading the shapefile of the sub-basin into FLO-2D, the boundary for the area was set by defining the boundary grid elements. The grid elements inside of the defined boundary were considered as the computational area in which the simulation will be run. The DTM of the area, which served as the elevation data for each grid, was imported in PTS format and was extrapolated into the model. A Manning’s roughness coeffcient of 0.03 was assigned to the grid elements of streams and rivers and 0.15 was used for the floodplain. Landcover was incorporated in the flood model to assign appropriate roughness and infiltration values in the base topography.
The next step was to allocate inflow nodes based on the locations of the outlets of the streams from the upper watershed. Outflow nodes were also allocated for the model. These outflow nodes show the locations where the water coming into the sub-basin is discharged. The water that will remain in the watershed will result to flooding on low lying areas. The outflow generated by the source sub-basin was used as inflow for the sub-basin area that it flows into.
Rainfall intensity-duration-frequency (RIDF) values of the Philippine Atmospheric, Geophysical and Astronomical Services Administration (PAGASA) were used for longer-term rainfall events (Kintanar, 1984; PAGASA, 2009), while a range of rainfall intensities from 30-70 mm/hr was used for short-lived downpour due to thunderstorms. Longer term rainfall events were used for the generation of flood maps that were used to pinpoint the intersection of streets and creeks corresponding to the MMDA list of flood prone areas. Once identified, higher resolution simulations of the flood-prone streets were conducted for thunderstorms with short-lived but intense rainfall.
After all the parameters were set, the model was run through FLO-2D GDS Pro with a simulation time of 1.5 hours. After running the flood map simulation in Flo-2D GDS Pro, Flo-2D Mapper Pro was used to read and generate the flow depth maps.
Based on observations of the location of MMDA and netizen accounts of flood-prone areas relative to the flood hazard maps of DOST-Project NOAH, an analysis is made on the root causes of street flooding in the Metro Manila. Examination of road profiles in flood prone areas and flood simulations for different rainfall scenarios were used in the study. Based on the analysis, solutions are proposed for further work by civil engineers in concerned government agencies.
3.1 Creeks and Street Intersection
When plotted in a map, the flood-prone areas list by MMDA becomes an online reference for Metro Manila commuters (Figure 2A). Overlaying the same list in a flood map shows they are related to streams and creeks (Figure 2B). Most of the places identified by MMDA are at the intersection of roads and flooded creeks captured in the 100-year rain return flood hazard map of DOST-Project NOAH. Those in the MMDA list that are not along intersections of creeks and highways are ponded areas depicted in the flood hazard map. An example of ponding is in Padre Burgos (City Hall) in Manila (Figure 2B).
Based on field investigation, out of the 23 places in the MMDA list of flood prone streets, 5 have bridges (see Appendix B – Table 2). These are Philcoa, R. Papa, C-5 Bagong Ilog, Osmeña-Skyway, and Don Bosco. In Philcoa, the street is raised above the bottom of the creek by about 3.8 meters with 1 rectangular, 2.37 x 4.4 m culvert perpendicular to the road and 2 circular, 1 x 1 m culverts parallel to the road and protruding into the stream’s channel. Along R. Papa, the street elevation is 1.38 m above the creek bottom and 0.4 m above the creek’s bank. There is a bridge above the intersecting creek to R. Papa Road, but with no culverts. The C-5 Bagong ilog site has a flowing stream with a bridge that spans 11 m. The street in this area is elevated about 3.92 m above the stream’s bottom and 1.14 m above the stream’s bank. Along Osmeña Skyway, a 22.3 m bridge spans above a flowing stream. The bottom of the stream and bank of the stream are 3.34 m and 2.74 m below street level, respectively. The last site in the list of MMDA flood prone streets with a bridge is found in Don Bosco, Makati. Here, there is a stream with its bottom and banks 4.5 m and 4 m below street level, respectively. A 4 x 2.5 m rectangular drainage structure positioned parallel to the street aids the flow of water into the stream.
In other areas without bridges, drainage networks were observed under the street, either as drainage laterals or culverts positioned perpendicular to the road (Table 2). Based on field investigation, eight places were reported not to have evident drainage networks, but this can be due to masking by the overlying concrete.
The profiles of all these roads in the MMDA flood prone list show that they are in topographic lows and not significantly elevated from the side of the road (see Appendix A – Figure 10). Based on profiles viewed in high resolution LIDAR topography, many of these roads are also not elevated relative to the top bank of the creek as measured from the road side. Along R. Papa, for example, the street that gets flooded have ”side of the road” profiles showing 3 channels, labeled in the figure as channel 1, channel 2 and channel 3. Apart from the road being at a topographic low, its elevation is below the top of the channel in each case. There is a clearance of about 2, 3 and 4 meters, respectively, before these channels can get filled by water. In EDSA-North Avenue, the part that gets flooded is the lowest portion of the street and follows the cross section of a 250-meter wide and 1-meter high channel. The road follows approximately this natural depression in the topography.
3.2 University of the Philippines – Old and New Roads
In the University of the Philippines, A. Roces Street and C.P. Garcia Avenue cross creeks (Figure 3). Both of these creeks are in the headwaters of a drainage network that leads to the Pasig River that drains toward Manila Bay. Profiles along these streets show that A. Roces Street is elevated from the creek by about 4 meters while C.P. Garcia is only elevated about 1.25 m from the bank of the creek (Figure 4A and 4C). In the case of C.P. Garcia there appears to be a bigger channel, where the road is lower in elevation than the bank of the creek (Figure 4C). These two streets have two culverts each with a diameter of 1 meter (Figure 4C and 4D).
The road along A. Roces is sufficiently elevated such that it does not flood even if the creek swells (Figure 10A). Brief but heavy rains delivered by thunderstorms can inundate the floodplains of the creek but never to street level. Even the peak floods during Tropical Storm Ondoy (international name Ketsana), a fatal flood event in 2009, did not inundate A. Roces Street. The creek in this part of the campus is uninhabited and is open ground with a lagoon and forest on the west side and an open field with trees on the east side.
In contrast, C.P. Garcia is known to flood even during short-lived but torrential thunderstorms, rendering the road temporarily impassable each time. Informal settlers living in the area beside the creek describe floods along this street to In contrast, C.P. Garcia is known to flood even during short-lived but torrential thunderstorms, rendering the road temporarily impassable each time. Informal settlers living in the area beside the creek describe floods along this street to have reached chest level (approximately 1 m high) during the peak of Ondoy. They also describe waist high floods during brief thunderstorms with heavy downpour.
3.2 Thunderstorm Flash Flood Scenarios
The MMDA list of flood-prone areas was compared to the flood depth and hazard maps simulated using 100-year rain return rainfall data from PAGASA. The same rain return period scenario was utilized for the analysis of inundation along A. Roces Street and C.P. Garcia Avenue.
The 100-year rain return period flood scenarios require simulations that reach up to 24 hours and generate flood maps that have high flood depths. Simulations for short-duration thunderstorm events up to 1 hour with intensities of 30-70 mm per hour show that flooding also takes place along roads in the areas listed by MMDA as flood prone. These flash floods are not as hazardous as those delivered by Habagat or storms but can cause nuisance traffic because they inundate up to tire level rendering streets impassable. Out of 23 in the list provided by MMDA, four are shown as examples: Bayani Road in Taguig City; Victory avenue in Quezon Boulevard; R. Papa Street in Manila; and S.M. North Avenue (Figure 5).
Bayani Road in Taguig starts to flood knee deep along the road at 40 mm/hr rainfall and can flood up to waist high with an hour of rainfall at intensity of 70 mm/hr. R. Papa Street begins to flood gutter deep across three sections of the street with 50 mm/hr of rain precipitation over a period of an hour and reaches half tire deep with a rain intensity of 70 mm/hr. Victory Avenue starts to flood up to the gutter at rain rates of 40 mm/hr, reaches knee deep at 60 mm/hr and floods tire deep at 70 mm/hr along the center of the road. The front part of SM North along EDSA begins to flood at 40 mm/hr and can reach up to tire high floods for 1 hour of rainfall with intensities of 60-70 mm/hr. Snapshots of simulations for the other 18 MMDA flood prone areas are provided as supplementary files in Appendix C (Figures 11-14).
The problem of floods in Metro Manila can be broadly classified as those that endanger people’s lives and floods that are considered a nuisance due to traffic jams they create. In both cases, however, it is clear that not all of Metro Manila is flood prone. There are places that are not hazardous to floods and are still suitable for development, while traffic-generating floods are limited to specified areas and have a possible relatively low-cost solution.
The MMDA list of 23 flood prone areas that cause traffic are observed mostly at intersections of streets and creeks and in some places are related to ponding of water such as in Padre Burgos in Manila. In flood prone streets that intersect with creeks, floods appear to occur because they are in natural topographic lows. Although the creek itself is the lowest point in the locality, the streets themselves are still in topographic depressions and not elevated enough relative to the channel as seen in profiles of the road side (Figure 10).
In the case of creeks in the UP Diliman Campus, two contrasting scenarios during severe weather events and short-lived thunderstorms are seen (Figure 3). A. Roces Street, built in the early 1990s, which is elevated by 4 meters above the top of the creek’s channel never floods (Figure 4A and C). C.P. Garcia Street on the other hand, built in the late 1970s and early 80s and constructed only 1.25 meters higher than the top of the creek’s channel, gets inundated during severe weather (i.e. tropical cyclones and southwest monsoon) as well as short-lived thunderstorm events (Figure 4B and D). It is also notable that the profile of the landscape beside the road has another wider channel apart from the creek’s channel that has a bank level that is higher in elevation relative to the road profile (Figure 4B).
These two contrasting examples of how roads are built relative to the elevation of the creek and the landscape demonstrate the difference on road design and the manner by which roads get flooded or not flooded. Roads that follow topographic lows, even if they are above the creek’s bank, should have enough clearance or freeboard to accommodate creek swelling during heavy downpour (Figure 6). Although small streams may oftentimes be dry or nearly dry especially during summer, they too have floodplains and can swell to overtop its banks. Design and construction of roads must therefore take into consideration swelling of the creek in the same manner that roads that cross large rivers take into account floods that overtop riverbanks.
Flood control in the Philippines follows technical standards based on design flood frequency expressed by return period (i.e., the probability, expressed in years, where a flood of a target size/magnitude is likely to occur). The return period is determined based on the size of catchment area, the degree of importance of the proposed project area and the economic viability of the project. Thus, it is necessary to determine the design flood discharge corresponding to the design flood frequency of the river and the funds needed for the implementation of the proposed improvement works and the expected benefits (JICA, 2003). Ideally, calculated discharges are obtained based on runoff analysis using available rainfall data. But in cases where there are sufficient past annual maximum flood data on the project site, it is more convenient to analyze these flood data compared with rainfall data. For example, there are available annual maximum flood discharge data in a 30-year period, the largest among these data is approximately the flood discharge for a 30-year to 60-year return period (JICA, 2003). However, rainfall data is normally the basis of determining the return-period of flood, as these are the only readily available data compared with discharge. For catchment areas below 20 km2, a Rainfall Intensity Duration Frequency Curve (RIDF) is utilized in calculating the discharge using the Rational Formula Method (Kuichling, 1889; Viessman and Lewis, 1995) (Eq. 3).
where Q is the discharge in m3/s, c is the runoff coefficient based on land cover, i is the rainfall intensity in m/s and A is the drainage area in m2.
In the design of the culverts and lateral drains found in the MDDA list of flood prone areas, it is comprehensible that they were not appropriate for unusually big floods that have plagued Metro Manila in the past years, such as the Ondoy, and the Habagat 2012, 2013 (Lagmay et al., 2014) and 2014 flood events. However, for short-lived thunderstorms with rainfall delivery of ≥30 mm/hr, it is boggling.
A computation of the theoretical discharge using the Rational Method was made for the watershed that drains into the creek that intersects the road along Philcoa, one of the places in the list of the MMDA flood prone areas (Figure 3). When calculated, the discharge expected along Philcoa for rainfall delivered over the span of 1 hour in a watershed area of 1,621,380 m2 and with intensity of 70 mm/hr is 29.95 m3/s. This value is still below the computed 87 m3/sdischarge capacity of the Philcoa culvert, derived using the Manning’s equation (Gauckler, 1867; Manning, 1891; Chanson, 2004) (Eq. 4).
where A is the inner area of the culvert in m2, n is the Manning’s coefficient of roughness, R is the hydraulic radius (area over the wetted perimeter) in meters and S is the slope of the stream bed at the culvert. Concrete box culverts can have a Manning’s n of 0.012 to 0.015. Since a higher roughness value will lead to a more conservative computation for the discharge capacity, 0.015 was used. The slope was obtained from the high-resolution topography.
Despite the calculated capability of the rectangular culvert to discharge the theoretical waterflow (i.e. Rational Formula), field interviews in the Philcoa area reveal that floods along the road at Philcoa happens even during brief thunderstorms. Field interviews also reveal that the culvert never gets topped except during Ondoy when it was full to the brim. The analysis leads us to suspect that other factors play an important role in generating thunderstorm-related floods in the area.
To determine other possible factors that cause the ephemeral urban flood, 2-D flood simulations were conducted. A 1.5-hour (90 minutes) simulation with a 70 mm/hr rainfall distributed over the first 60 minutes of the simulation was used to observe flooding in the area. This revealed that street flooding along Philcoa begins at 33 minutes (Figure 7a) and is maximum at 39 minutes (Figure 7b) from the start of the simulation. After this, flooding slowly decreases in magnitude. Only at the end of the simulation at 90 minutes (Figure 7d) has the flood drained fully into the creek. The peak discharge at the culvert’s inlet happens at 57 minutes (Figure 7c) from the start of the simulation, in which we can see that the flood along the road has already started to wane. This means that floods along the road start earlier than peak discharge along the culvert.
The hydrograph (Figure 8) also shows a low point after peak flow at 1 hour 20 minutes before it begins to rise again. The initial rise is attributed to surface water nearby draining into the culvert while the succeeding rise in the discharge is attributed to flow contributed from farther reaches of the watershed.
One limitation of the study is the absence of information of subsurface lateral drainage in the Philcoa area. However, field work allowed insights into the possible subsurface street drainage. In the road flood simulation, we have made sure that a 1 m x 1 m exposed drain directly connects with the Philcoa creek. This mimics unexposed lateral drainage that allows flow directly into the creek.
The 2D flood simulation tells us that even if the design of the culvert does have the capacity to discharge the theoretical discharge rate fed by surface flow from heavy downpour, significant flooding still occurs for at least 24 to 30 minutes, after which it subsides until fully draining after another 30 minutes. This then causes costly urban traffic. The high rate of surface flow into the topographic low in the Philcoa area is faster than the rate by which surface flow drains into the creek. Since the creek never becomes full during short-lived thunderstorms, even with some garbage, it is really not the culvert’s design but the road design built following the topographic low that is the problem.
A possible solution may be derived from the example of A. Roces Ave. in UP Diliman Campus where the road is significantly elevated relative to the creek’s banks. An alternative solution is to construct a hollow portion under the street big enough to accommodate the volume of water, which is equivalent to the area covered by the flood multiplied by the height of inundation, which for the case in the Philcoa area is only up to knee-high. The street to be constructed must have steel mesh drains all over to accommodate the high rain rates that accumulates almost instantaneously in the topographically low-lying area. The subsurface water retention area should ideally be designed to flow into the creek if the capacity of the culvert crossing the street can accommodate high discharge rates such as the case in Philcoa. Constructing the retention basin below the street, if structurally feasible, avoids right-of-way problems.
The proposed solution of creating a basin below the street, which drains directly into the creek is simulated for the road along Philcoa. Since the Flo-2D model can’t simulate flow below the street, an alternative model was generated. Strips parallel to the street were excavated in the DTM and designed to slope towards the Philcoa creek. The result of the simulation shows all the flood water confined in the excavated strips with no flood on the road. Since the excavated strips are all underground, it means that the top of the strip or road surface will not be flooded as well (Figure 9). This type of simulation can be used to determine the appropriate surface infiltration of the street, and the depth and slope of the subsurface retention basin that is needed in order for water not to pond during short-lived thunderstorm events.
Although one example of discharge analysis was presented, the findings could be applied to other places in the list of 23 flood prone areas in Metro Manila. They are applicable to other creek and street intersections with culverts (e.g. Don Bosco) since they have similar geometry for computing discharge. For ponded areas like in Padre Burgos in Manila and at EDSA Pasong Tamo, retention basins can also be made under the street and pumped out after the thunderstorm. However, at EDSA Pasong Tamo, this type of intervention may be difficult as there is an overpass in the area. High volume pumps are necessary for this area to take out floodwaters into an adjacent retention basin.
Metro Manila’s floods are compounded by many factors including encroachment of concrete surfaces, densification of buildings and residential areas, silting of riverbeds and canals, obstruction of waterways by informal settlers, clogging of floodways by garbage, narrowing of rivers due to development on floodplains, draining and filling in of small rivers forcing more water into fewer channels, forest degradation, and reclamation of coastal land. Furthermore, humans have altered the landscape in the metropolis which has grown rapidly but with poorly planned urbanization. Since the 1970s, people have migrated from rural areas to Metro Manila increasing the population from 4.9 million residents in 1975 to more than 11 million today. A survey by the National Housing Authority showed that by the early 1980s, a quarter of Metro Manila residents were informal settlers living in crowded shantytowns many along waterways. Further complicating the problem is ground subsidence. From 1978 to 2000, parts of Metro Manila sank by an amount ranging from 16 centimeters to 1.46 meters. The probable causes of subsidence are excessive groundwater extraction, soil compaction and tectonic movement, though more research is needed to fully determine the primary causes (Lagmay et al., 2010).
Proposed solutions to street floods in many of the places in the flood-prone list by MMDA include the elevation of roads and construction of a retention basin under the street big enough to accommodate the volume of water. The size of the retention basin is equivalent to the simulated area covered by the flood along the street multiplied by the height of inundation. The street above the proposed retention basin must have enough infiltration capacity to accommodate the rapidly accumulated rainfall. The retention basin must also be designed to directly drain into the nearest stream channel. The latter solution or its equivalent may not completely address big floods spawned by severe weather events but may solve frequent street floods from short-lived thunderstorms. These proposed solutions may avert increased gas consumption and lost potential income from mammoth traffic jams during floods, which costs the Philippine economy 2.4 billion pesos a day, a figure that may balloon to 6 billion a day by the year 2030.
Chanson, H.: The Hydraulics of Open Channel Flow, Butterworth-Heinemann, Oxford, UK, 2nd edn., pp. 630, 2004.
Cox, W.: The evolving urban form: Manila, New Geography http://www.newgeography.com/content/002198-the-evolving-urban-form-manila, accessed 9 September 2013, 2011.
Gauckler, P.: Etudes Th´eoriques et Pratiques sur l’Ecoulement et le Mouvement des Eaux, Comptes Rendues de lA´ cade´mie des Sciences, 64, 818–822, 1867.
JICA: Manual on flood control planning, Project for the Enhancement of Capabilities in Flood Control and Sabo Engineering of the DPWH Department of Public Works and Highways, Japan International Cooperation Agency, 2003.
Kintanar, R.: Climate of the Philippines, Tech. rep., PAGASA, 1984.
Kuichling, E.: The relation between the rainfall and the discharge of sewers in populous districts, Transactions, American Society of Civil Engineers, 20, 1–56, 1889.
Lagmay, A., Rodolfo, R., and Bato, G.: The perfect storm: Floods devastate Manila, Earth, 55, 50–55, 2010.
Lagmay, A., Crisologo, I., Racoma, B., and David, C.: Volcanoes magnify Metro Manila’s southwest monsoon rains and lethal floods, Frontiers Earth Science, pp. 60–105, doi:10.3389/feart.2014.00036, 2014.
Manning, R.: On the flow of water in open channels and pipes, Transactions of the Institution of Civil Engineers of Ireland, 20, 818–822, 1891.
PAGASA: Member Report to the ESCAP/WMO Typhoon Committee 41st Session, Tech. rep., Philippine Atmospheric Geophysical and Astronomical Services, 2009.
Rodis, R.: Manila’s traffic jams cost $57 million a day, http://globalnation.inquirer.net/113269/manilas-traffic-jamscost-57-million-a-day, accessed 10 December 2014, 2014.
Viessman, Jr., W. and Lewis, G.: Introduction to hydrology, John Wiley and Sons, 2nd edn., pp. 311, 1995.
Zoleta-Nantes, D.: Flood Hazards in Metro Manila: Recognizing Commonalities, Di_erences, and Courses of Action, Social Science Diliman, 1, 60–105, 2000.
Road against roadside profiles
Tabulation of field characteristics of MMDA flood prone areas
Simulations of other MMDA flood prone areas
Source: Typhoon Ruby (Hagupit) Update (NOAH DOST)
Street floods in Metro Manila and possible solutions