Hydrological Response of the Pampanga River Basin in the Philippines to Intense Tropical Cyclone Rainfall

Rhonalyn V. Macalalad Institute of Environmental Science and Meteorology, University of the Philippines Diliman, Quezon City, Philippines
Philippine Atmospheric, Geophysical and Astronomical Services Administration, Quezon City, Philippines

Search for other papers by Rhonalyn V. Macalalad in
Current site
Google Scholar
PubMed
Close
,
Roy A. Badilla Philippine Atmospheric, Geophysical and Astronomical Services Administration, Quezon City, Philippines

Search for other papers by Roy A. Badilla in
Current site
Google Scholar
PubMed
Close
,
Olivia C. Cabrera Institute of Environmental Science and Meteorology, University of the Philippines Diliman, Quezon City, Philippines

Search for other papers by Olivia C. Cabrera in
Current site
Google Scholar
PubMed
Close
, and
Gerry Bagtasa Institute of Environmental Science and Meteorology, University of the Philippines Diliman, Quezon City, Philippines

Search for other papers by Gerry Bagtasa in
Current site
Google Scholar
PubMed
Close
https://orcid.org/0000-0002-5433-7122
Full access

Abstract

The Philippines is frequently affected by tropical cyclones (TCs), and understanding the flood response of the Pampanga River basin (PRB) from TC-induced rain is needed in effective disaster risk management. As large uncertainties remain in TC rain forecasting, we propose a simple checklist method for flood forecasting of the PRB that depends on the general TC track, season, and accumulated rainfall. To this end, flood events were selected based on the alert, alarm, and critical river height levels established by the Philippine Atmospheric, Geophysical and Astronomical Services Administration (PAGASA). Results show that all flood events in the PRB were induced by TCs. All intense TCs that directly traversed the PRB resulted in critical-level floods. These TCs also had the shortest flood onset of 7–27 h from alert to critical level. Flooding from distant landfalling TCs, on the other hand, are dependent on season. TCs traversing north (south) of the PRB induced flooding only during the southwest (northeast) monsoon season. These TCs can raise water levels from alert to critical in 11–48 h. Remote precipitation from non-landfalling TCs can also induce critical-level flooding but with a longer onset time of 59 h. These results indicate that a simple checklist method can serve as a useful tool for flood forecasting in regions with limited data and forecasting resources.

© 2021 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author: Gerry Bagtasa, gbagtasa@iesm.upd.edu.ph

Abstract

The Philippines is frequently affected by tropical cyclones (TCs), and understanding the flood response of the Pampanga River basin (PRB) from TC-induced rain is needed in effective disaster risk management. As large uncertainties remain in TC rain forecasting, we propose a simple checklist method for flood forecasting of the PRB that depends on the general TC track, season, and accumulated rainfall. To this end, flood events were selected based on the alert, alarm, and critical river height levels established by the Philippine Atmospheric, Geophysical and Astronomical Services Administration (PAGASA). Results show that all flood events in the PRB were induced by TCs. All intense TCs that directly traversed the PRB resulted in critical-level floods. These TCs also had the shortest flood onset of 7–27 h from alert to critical level. Flooding from distant landfalling TCs, on the other hand, are dependent on season. TCs traversing north (south) of the PRB induced flooding only during the southwest (northeast) monsoon season. These TCs can raise water levels from alert to critical in 11–48 h. Remote precipitation from non-landfalling TCs can also induce critical-level flooding but with a longer onset time of 59 h. These results indicate that a simple checklist method can serve as a useful tool for flood forecasting in regions with limited data and forecasting resources.

© 2021 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author: Gerry Bagtasa, gbagtasa@iesm.upd.edu.ph

1. Introduction

Rainfall in parts of the Philippines can reach in excess of 4000 mm yr−1 The interaction of seasonal monsoon wind with land (i.e., orographic, convective) gives rise to a distinctive west and east coast rainfall in the boreal summer (southwest monsoon) and winter (northeast monsoon) seasons, respectively (Akasaka et al. 2007; Bagtasa 2020b). In addition to monsoonal rainfall, extreme precipitation from tropical cyclones (TCs) also contributes significantly to rainfall in the Philippines (Bagtasa 2017). High runoffs from intense rainfall coupled with a complex topography and small drainage areas expose a large portion of the population to a high risk of flooding. In the present study, we focus on the hydrological response characteristics of the Pampanga River basin (PRB) in the Philippines. The PRB is the Philippines’ fourth largest river basin located in the central plain of Luzon island in northern Philippines. The PRB lies within the longest contiguous lowland plain in the Philippines where a third of the country’s rice production is from (DENR 2019), hence, its importance to the country’s food security. This area is often locally referred to as the “rice granary of the Philippines.” Past flooding events in the PRB affect about 60% of the region’s 11.2 million population (PSA 2011; Shrestha et al. 2016). The region has a thriving economy that contributes to 9.2% of the Philippines’ GDP (NEDA 2018). Due to the continuous rapid economic development in the basin and the projected impacts of climate change, annual flood damages are also likely to increase correspondingly.

The PRB covers the provinces of Bulacan, Pampanga, and Nueva Ecija, which often suffer from periodic extensive flood damage. Previous studies (JICA 2011; Shrestha et al. 2016) show the PRB experiences an average of at least one significant flooding annually. The main cause of the reported frequent and large-scale flooding events in the basin is attributed to intense rainfall that leads to exceedance in river flow capacity. Rainy season in central Luzon coincides with the western North Pacific (WNP) summer southwest monsoon which starts in June and ends in October (Wang 2002; Akasaka et al. 2007). Previous PRB flooding events mainly occurred in the wettest months from July to September. In addition, an average of 1.7 TCs crosses the basin or nearby regions every year (Okazumi et al. 2014). The flood events that occurred in August 1960 due to the effects of Typhoon (TY) Agnes (Ferraris 1971), July 1962 due to TY Kate, and the “Great Philippine flood of July 1972” in July of 1972 due to TY Gloring (international name: Rita) were some of the most devastating events that affected the PRB (Ferraris 1971; Gordon 1973). In the recent two decades, three consecutive flood events that occurred within the river basin in November of 2004 had left large-scale damages affecting a large portion of the population and properties. The case of Tropical Storm (TS) Ondoy (Ketsana) and TY Pepeng (Parma) in September 2009 led to human casualties which included 56 dead, 3 injured, 7 missing, and a total of 882 houses damaged in the region (Yumul et al. 2013). TY Pedring (Nesat) of September 2011 resulted in 55 fatalities due to drowning and 43 032 houses damaged (NDRRMC 2011). Furthermore, climate change also poses a risk in future occurrences of flooding in the PRB. Recent high-resolution multimodel ensemble simulations show a tendency of future climate projection toward warmer and wetter conditions in the island of Luzon (Villafuerte et al. 2020). This will make the increase in future flooding a virtual certainty, especially in the upland areas of the basin (Jaranilla-Sanchez et al. 2013; Tolentino et al. 2016). This is in addition to the increased frequency and further strengthening of TCs in a warmer climate (Emanuel 2005; Webster et al. 2005).

Despite the severity of the aforementioned flooding events both in the past and in future climate change projections, limited studies have been undertaken on the hydrometeorological behavior of this catchment. Focus is placed mainly on recovery efforts and/or damage assessment (Okazumi et al. 2014; Shrestha et al. 2016) rather than further understanding how the watershed responds to meteorological forcings. The response of a watershed to precipitation or evapotranspiration is nonlinear and cannot be predicted with certainty due to the inherent complexities of hydrological systems (Gebrehiwot et al. 2011). Furthermore, the knowledge of processes governing a basin’s hydrological response is still generally limited (Penna et al. 2016) and basins exhibit heterogeneous and complex rainfall–runoff processes in different hydroclimatic regimes and at different scales (McDonnell et al. 2007). Still, analysis of past events could bring great benefit for future prediction efforts and disaster management in the Philippines, particularly for the PRB. The present study aims to characterize the PRB hydrological response to intense rainfall by using gauge data located within the basin to draw general conclusions on the meteorological conditions under which flooding is generated. This paper is presented as follows. Section 2 describes the PRB gauge station data used and the method in the determination of flooding events. Section 3 discusses the events when PRB water level exceeded PAGASA’s three-level threshold and the coincident heavy rain events. Finally, section 4 summarizes the study and presents the conclusions.

2. Data and methodology

a. Study area

The PRB is the fourth-largest river basin in the Philippines located in Central Luzon in the largest island of the Philippines. Figure 1 shows the map of the PRB. It has a catchment area of 10 434 km2 including the allied basin of Guagua River. It covers the provinces of Nueva Ecija, parts of Bulacan, Tarlac, Quezon, and almost the entirety of Pampanga. Highland areas enclosing the basin includes the Sierra Madre to the east, Caraballo Mountains to the north, and parts of the Zambales Mountains where Mt. Pinatubo is located to the west. Of these boundaries, the Zambales Mountains to the west are characterized by ultramafic rocks with some volcanic components (Schopka et al. 2011). The Sierra Madre range to the east comprise more of dissected and eroded terrain (JICA 2011). The central part of the study area is called the Central Luzon Plain due to its relatively flat terrain. It is an alluvial plain with a ground elevation below 200 m and slope less than 3% or 0.0006 that spreads over the lower and middle reaches of the PRB (JICA 2011; Lee et al. 2013). The peak elevation of the basin is located at the northeast boundary with the Sierra Madre range at 1885 m above mean sea level. The lowest elevation, almost equal to mean sea level, extends widely around the river mouth of the PRB.

Fig. 1.
Fig. 1.

PRB topography with provincial boundaries and location of PAGASA gauging stations. Blue triangles are rainfall stations while the red triangles are combined rainfall and water level stations.

Citation: Journal of Hydrometeorology 22, 4; 10.1175/JHM-D-20-0184.1

The main river of the PRB, the Pampanga River, has a total length of about 260 km. Its tributaries include the Penaranda and the Coronel–Santor Rivers to the east, and the Chico River to the northwest. On the eastern side of the river adjacent to Mount Arayat is the Candaba swamp covering an area of some 250 km2. It absorbs most of the flood flows coming from the eastern portion of the basin and the overflowing water of the Pampanga River via the Cabiao Floodway. This area is regarded as a retarding basin during the rainy season (Shrestha et al. 2016). In the downstream reaches from Mt. Arayat, the Pampanga River has a continuous dike along the west bank, which separates the Guagua River basin from the Pampanga River. The Angat River joins the Pampanga River at Calumpit in Bulacan through the Bagbag River. In the lower reaches of the basin, the Pampanga River crosses the Pampanga delta where numerous settlements with fertile farmlands were developed in the northern portions and fishponds in the southern portions. The river eventually empties into Manila Bay.

b. Data

Data from 10 ground-based rainfall gauge stations from 2013 to 2018 are utilized in this study. Six of the 10 stations are combined with a water level gauge positioned across the river basin. The PRB map in Fig. 1 shows the location of the rain/water level gauges. The data at every station are transmitted in near–real time to the Philippine Atmospheric, Geophysical and Astronomical Services Administration (PAGASA, national weather agency) main office and subcenter office through a telemetry system. Table 1 shows the station names, observed parameters, and coordinates of gauging stations within the PRB. Stations in the downstream area of Arayat station were not included in the study. TC tracks are from the Tropical Cyclone Guidance System (TCGS) of PAGASA.

Table 1.

Name, type, and location of gauging stations.

Table 1.

c. Methodology

In this study, we analyzed the response of the Pampanga River to heavy rainfall events by using the water level observed in Arayat station. Arayat station is the most downstream station where water from the three major subbasins (Angat, Pasac, and the main Pampanga basin: see Fig. 1) meet, and is not affected by rising tides from Manila Bay to the south. Consequently, the rising river level in this station signifies flooding around the region and downstream areas. Hence, it was selected to characterize the river response.

Flood events were defined by creating a continuous plot of basinwide rainfall together with water level data from the Arayat station from 2013 to 2018. The periods where the observed water level reached the PAGASA-established alert level of 5 m, alarm level of 6 m, and a critical level of 8.5-m river height were selected. The water level is at the alert, alarm, and critical levels, when the river is 40%, 60%, and 100% full, respectively. Each rainfall event that caused a threshold-breaching rise in water level was analyzed. For mean basinwide rainfall in the PRB, this was estimated using the Thiessen polygon method (Fiedler 2003) where rainfall in each station upstream of Arayat station was included in the calculation. Due to the limited rainfall observation points, it is assumed that areas in the watershed receive the same amount of rainfall as the nearest gauge. The observed rainfall depth is applied to any point at a distance halfway to the next station in any direction. In this study, the Thiessen polygon method was applied by creating subbasins that were delineated using Hydrologic Engineering Center’s Geospatial Hydrologic Modeling (HEC-GeoHMS) utilizing interferometric synthetic aperture radar (IFSAR) (5-m resolution) digital elevation model data from the Philippine National Mapping and Resource Information Agency (NAMRIA). After this, the percentage weight of rainfall that contributes to each subbasin was determined. The subbasin mean rainfall was then calculated followed by calculating the areal mean rainfall.

To determine the rate of increase in river water level, we first observe the time it takes for the water level at Arayat station to rise from the initial level, or the level prior to intense rainfall, to the alert level. In addition, we also determined the time between water rising from alert level to alarm level. The average rate of flood onset was computed by getting the ratio of the net change in water level and the elapsed time. The accumulated rainfall for the onset was also identified by summing the rainfall amount from the initial water level until the water level reaches the alert level. The same procedure was done for the accumulated rain from alert to alarm level. The accumulated rainfall that caused the water to rise from the initial level to the assessed level was calculated by averaging all observed rainfall in the upstream stations. We also calculated the lag time or the time from peak rain intensity to the peak water level (Archer and Fowler 2018).

To measure the influence of antecedent watershed soil moisture, the antecedent precipitation index (API) is computed given by the equation APId=Pd+kPd1+k2Pd2++KnPdn, where APId is the antecedent precipitation index for the day d, k is an empirical decay constant set at 0.9, Pd is the rainfall for the day d, and n is the number of days prior to day d set at n = 30 (Linsley et al. 1975). River discharge is also calculated in this study based on the observed water level height using a rating curve derived by PAGASA given by Q = a(Hb)c, where Q is the discharge/river flow (m3 s−1), H is the water level in meters, and a, b, and c are dimensionless coefficients unique to specific river channels. For the Arayat station, the most recent rating curve equation from a 2011 survey of PAGASA is Q = 14.86(H + 1.05)2 and Q = 21.05(H − 0.55)2 for water level below and above 8.93 m, respectively. These rating curves are calibrated and validated with flood events of 2011 and 2012 (Shrestha et al. 2016).

3. Results and discussion

a. Heavy rain events

Figure 2 shows the hydrograph of Arayat station covering the period of 2013–18. The hydrograph illustrates the observed rainfall and water level of the river. In addition, the red, orange, and yellow lines show the critical, alarm, and alert levels, respectively. The river height is used to determine the potential of the river to cause flooding in the downstream regions and/or low-lying areas in the basin. In the dry months from January to May, water level observed at the Arayat station has a mean value of 0.65 m. The start of the rainy season in Central Luzon corresponds with the onset of the southwest monsoon season, which begins at the end of May to early June (Akasaka et al. 2007; Moron et al. 2009). It is apparent that rainfall and water level starts to increase during this time. Rainfall in the basin peaks from July to September, then gradually decreases until the month of December. These peak rainfall months also coincide with the most active TC season in the WNP region. To characterize the hydrological response of the PRB, we first present the events that led the water level to breach the critical, alarm, and alert levels. In the 6-yr study period, there are eight instances where the water level breached the critical level. All of these instances occurred with at least one TC affecting the Philippines. Each of these critical-level status events is highlighted in Fig. 2 and the corresponding TCs and their tracks are shown above and below the hydrograph.

Fig. 2.
Fig. 2.

Hydrograph data from 2013 to 2018 in PRB showing identified rain/flood events based on exceedance of alert (yellow line), alarm (orange), and critical level (red line). Black bars denote basin mean rainfall and dark blue lines depict water level while pink boxes highlight critical-level events with tracks of coincident tropical cyclones that entered the Philippine Area of Responsibility. Tracks are from the TCGS of PAGASA.

Citation: Journal of Hydrometeorology 22, 4; 10.1175/JHM-D-20-0184.1

Almost all critical flood events are associated with landfalling TCs, which carry intense rainfall and are termed as events hereinafter. TCs do not necessarily traverse the PRB directly to induce flooding. Of the eight TCs, six are classified as typhoons and two as tropical storms (TS). One of the TS, TS Inday (international name: Ampil) in 2018, coexisted with another tropical depression (TD Josie), both of which did not make landfall in any Philippine landmass during its associated rainfall event. Nevertheless, the two TCs were able to raise the water level at Arayat station above the critical level. Table 2 summarizes the local and international names of the TC events, their classification, dates of occurrence, and time on Philippine landmass.

Table 2.

Description of the eight TC events that led to increase water to critical level, their local (and international) names, categories, date of occurrences, and time on land.

Table 2.

The alarm level is reached when the river elevation is between 6 and 8.49 m. In the study period, 10 alarm level events are identified. Similar to the critical-level events, all alarm level events also occurred with at least one TC affecting the Philippines. Figure 3a lists the TC names, category, and dates of occurrence of all alarm level events. Two TCs were TY, six were TS, and one supertyphoon (STY). We note that the TY and STY alarm events [TY Labuyo (Utor), TY Yolanda (Haiyan), and STY Lawin (Haima)] brought significant impact to the basin and increased the river level close to the critical level. Figure 3b shows the list of 11 TCs that coincide with the increase in water level to the 5-m alert level mark. Of the 11 TCs, seven were classified as TY, three were TS, and one TD.

Fig. 3.
Fig. 3.

Tracks, local (and international) names, classification, and dates of the TCs that caused alert and alarm levels in PRB. Tracks are from the TCGS of PAGASA.

Citation: Journal of Hydrometeorology 22, 4; 10.1175/JHM-D-20-0184.1

Monsoon rains from the onset of the rainy season can raise the water level to approximately 2-m height. However, it is observed that TC-associated heavy precipitation triggered all alert, alarm, and critical-level events of the PRB. It is apparent in the hydrograph (Fig. 2) that water height exceeding the critical and alarm levels are typically seen as abrupt peaks rather than prolonged high water level status. Intermittent but intense TC rainfall explains why flood events are seen in such short-term peaks. This is consistent with rainfall in the adjacent Metropolitan Manila where abnormally wet years are determined by the intermittent effects of TCs rather than seasonal monsoon strength (Bagtasa 2020a). For the critical water level events, seven out of the eight TC events made landfall in Luzon. While it is known that TC rainfall is most intense around the inner core and weakens toward the outer rainbands (Jiang et al. 2013), distant non-landfalling TCs interacting with monsoon flow has been shown to induce intense remote TC precipitation in the northwestern regions of the Philippines (Cayanan et al. 2011; Bagtasa 2017). In fact, more than 90% of heavy rainfall events along western Luzon, to which a large portion of PRB belongs, are due to these non-landfalling TCs (Bagtasa 2019). Nevertheless, almost all critical-level events are caused by landfalling TCs.

A typical condition of a summer remote TC precipitation effect is a TC of at least TS category located to the northeast of Luzon and a TD embedded in a deep Asian monsoon trough in the northern regions of the South China Sea. Such TC(s) induces strong zonal moisture flux to western Luzon that results in heavy rainfall events during the southwest monsoon season (Bagtasa 2019). The non-landfalling TC event that produced heavy rainfall in July of 2018 at the PRB was caused by two interacting TCs, TS Inday and TD Josie. TS Inday was located to the northeast of the Philippines while TD Josie was located in the northern South China Sea. The locations of TS Inday and TD Josie conform with the synoptic condition of a typical TC–monsoon interaction that leads to heavy precipitation events in western Luzon.

Of the seven landfalling TCs, four TCs directly traversed the basin (TY Santi, TS Kabayan, TY Lando, and TY Karen), two crossed to the far north (TY Luis and TY Ompong), and one to the south of the basin (TY Nona). The two TCs that traversed north of the basin occurred in summer and the one TC that traversed south of the basin occurred in winter. We also note that all the distant landfalling TCs (not traversing the PRB) that resulted in critical-level events were TY categories and at least TS category below critical-level floods. Other distant landfalling TCs moving north and south of the PRB in northeast and southwest monsoon season, respectively, did not lead to flooding events. Moisture transport to TCs from distant ocean sources makes up a large proportion of a TCs’ total precipitable water (Kudo et al. 2014). This intrusion of moisture evaporated from remote sources to a TC’s center along the prevailing monsoon flow explains the dependence of TC track location in relation to the PRB and monsoon season. Moisture flux is highest along the confluence of the prevailing monsoon wind and TC circulation (not shown). Moisture flows from the southwest of a TC in summer and from the east to northeast of a TC in winter. As a result, convective outer TC rainbands form along these monsoon–TC wind confluence zones. Rain from these feeder bands occurs over the PRB region when a TC is located to the north of the basin during summer or to the south of the basin during the winter monsoon season.

Below critical-level events, more non-landfalling TCs led to alarm status. Half of the 10 TCs that raised the water level to alarm are non-landfalling TCs. Also, 80% of the alarm events were early season (July–September; southwest monsoon) TCs. This is almost similar to the case of alert level status wherein 7 out of the 11 TC events are non-landfalling. Furthermore, 72% (8 out of 11) of the events were also early-season TCs. All TCs during the southwest monsoon season that contributed to raising of water level traversed north of the PRB except TY Glenda, which tracked just along the southern periphery of the basin. In contrast, almost all TC events during the northeast monsoon season tracked south of the PRB. With some even traversing tracks as far south as the central Visayas region (e.g., TY Yolanda and TS Marce). Similar to the case of TY Glenda, TY Rosita traversed the immediate north of the PRB during the northeast monsoon, which led to an alert level event. Moreover, late-season (October–December) TCs tend to move in a more westward direction (Bagtasa 2017), which have a higher chance of making landfall and move across the PRB. Thus, the high number of late-season TCs leading to critical events even after the end of the region’s rainy season. Table 3 summarizes the TC events according to track location and monsoon season.

Table 3.

Tropical cyclones and their position relative to PRB’s location and season they occurred (asterisks indicate non-landfalling TCs). Note: STY category by PAGASA operationalized from 2015.

Table 3.

The following are the characteristics of rainfall events that led to the raising of river height: 1) Heavy precipitation events induced by TCs cause river swelling. 2) All intense, landfalling TCs directly traversing the PRB result in critical-level status. In the study period, all basin-crossing TCs occurred during the northeast monsoon season. 3) Intense landfalling TYs outside of the PRB can also result in critical-level status provided these TYs traverse north of the PRB during southwest monsoon season and south of the PRB during the northeast monsoon season. 4) In the case of alarm and alert events, most events occurred during the southwest monsoon season. This is because 50.0% of alarm and 71.4% of alert level events were caused by non-landfalling TCs, which only occur during the summer monsoon months. Other alarm and alert events follow (3) but can be at least TS category.

b. Hydrological response

Two metrics are used in characterizing the hydrological response to intense rainfall: the speed of onset of flooding (slope in hydrograph) and the speed of response or lag time from peak rainfall of the river at Arayat station. These metrics demonstrate how the gauge information based on river level measurements at 1-h interval behaves upon reaching the critical level. The hydrographs of the TCs that induced flooding due to exceedance of the critical river level are shown in Fig. 4. Figure 5 shows the time intervals between the three water levels and the time duration between critical to peak levels for each of the critical events. Table 4 includes information on the peak water level, lag time, accumulated rainfall for every river level, and the API of each TC event. To determine the factors that influence lag time and flood onset, quantitative analysis of TC events is separated into the three TC types that led to critical-level events: 1) landfalling TCs that directly crossed the PRB, 2) landfalling TCs that did not cross the PRB, and 3) non-landfalling TCs. We note that focused is mainly on the critical-level events as these events have the most potential for disaster.

Fig. 4.
Fig. 4.

Hydrographs during the passage of TCs (a) TY Santi (Nari), (b) TY Luis (Kalmaegi), (c) TS Kabayan (Mujigae), (d) TY Lando (Koppu), (e) TY Nona (Melor), (f) TY Karen (Sarika), (g) the combined effect of TS Inday (Ampil) and TD Josie TY, and (h) Ompong (Mangkhut).

Citation: Journal of Hydrometeorology 22, 4; 10.1175/JHM-D-20-0184.1

Fig. 5.
Fig. 5.

Time interval between water levels from alert to alarm (yellow), from alarm to critical (orange), and from critical to peak level (red) in hours.

Citation: Journal of Hydrometeorology 22, 4; 10.1175/JHM-D-20-0184.1

Table 4.

Description of the eight TC events that led to increase water to critical level, their categories, and date of occurrence.

Table 4.

1) TCs directly crossing the PRB

Four TCs traversed the PRB that led to critical-level status events. TY Santi was a category 3 TY that pummeled through Central Luzon on 12 October 2013 and remained on land for 8 h. During its transit on land, the recorded water level at Arayat station reached 8.78 m. This resulted in flooding in communities downstream. TY Santi left 13 dead and affected 200 000 families with more than 53 040 houses left damaged (NDRRMC 2013). TS Kabayan had a fast lateral speed and traversed the PRB in only 6 h on 2 October 2015. Although its duration on land is short, it had a mean peak rainfall rate of 18.72 mm h−1 and an accumulated rainfall of 98 mm that raised the river height to alert level. TY Lando was an intense category 4 TY that crossed the PRB on 18 October 2015. It remained on land for 33 h (~20 h in PRB vicinity) as its track recurved northward along the western Luzon coast. Its recurving track yielded an accumulated rainfall of 148 mm in the PRB region that resulted in an abrupt increase of water level from alarm to critical status in 14 h with a peak river level of 10.03 m. It affected around 1 532 331 individuals with 16 casualties in Central Luzon (NDRRMC 2015). Last, TY Karen crossed the PRB on 16 October 2016 also as a category 4 TY. The water level in Arayat station went critical and peaked at 8.68 m the following day that resulted in flooding at low-lying downstream areas.

Among the four TCs that crossed the PRB, TS Kabayan and TY Karen had similar lag times of 33 and 34 h, respectively. Both these TCs have very similar tracks while traversing Luzon. The two TCs also had similar accumulated rainfall amounts of 98 mm for TS Kabayan and 95 mm for TY Karen to reach the alert level. Although it took only 3 h to increase from alert to alarm level, both had long alarm to critical-level onset times of 24 h for TS Kabayan and 16 h for TY Karen. The shorter onset time of TY Karen is mainly attributed to its intensity and higher rain rate.

TY Santi had the shortest lag time between peak rainfall to alert level in just 30 h. This is likely due to the very intense rainfall of up to 90 mm h−1 in the basin as shown in its hydrograph in Fig. 4a. Its straight westward track passing directly at the center of the basin has also brought intense rainfall in the upstream regions. Overall, the hydrographs of the basin-crossing TCs—TY Santi (Fig. 4a), TS Kabayan (Fig. 4c), TY Lando (Fig. 4d), and TY Karen (Fig. 4f)—show a sudden surge of increase in water level. This suggests that the change in height of water level is mainly driven by the TCs’ intense rainfall leading to a large amount of accumulated rainfall in the basin. These basin-crossing TCs took only 2–3 h for the water to rise from alert to alarm level, 5–24 h to rise from alarm to critical level, and remained for 11–33 h above the critical level.

2) Landfalling TCs outside the PRB

TY Lando has a similar response as TY Nona which tracked south of the basin. TY Nona was a category 4 TY that made landfall in Northern Samar (southeast of PRB) on 14 December 2015. It made several other landfalls in the islands of northern Visayas as it moved west-northwest toward southern Luzon for 30 h before exiting to the South China Sea on 17 December. It caused flooding ranging from 0.3 to 2.4 m in low-lying regions of the PRB which subsided after several days (NDRRMC 2015). Despite its peak rainfall reaching 90 mm h−1, similar to that of TY Santi, its southerly track resulted in a different river response from TY Santi. TY Lando and TY Nona both caused major inundation in Central Luzon in 2015 and had lag times of 43 and 47 h, respectively. TY Nona brought heavier rainfall as compared to TY Lando but TY Nona’s response time was relatively longer. This difference is attributed to the associated earlier precipitation recorded upstream. Based on the PRB Flood Forecasting and Warning Center (PRBFFWC) postflood reports (PRBFFWC 2015), TY Lando’s effect on the PRB started with continuous heavy to intense rainfall for almost 24 h that was also observed in an upstream station at Gabaldon (~72 km northeast of Arayat Station). In contrast to TY Nona, although the mean basinwide observed rainfall was high, precipitation in upstream stations were relatively lower due to its southerly track. The hydrograph of TY Nona (Fig. 4e) still shows an abrupt increase in water level comparable to the TCs that directly crossed the PRB.

In addition to TY Nona, two other landfalling TC events–TY Luis and TY Ompong–did not cross the PRB but led to critical events. Both TY Luis and TY Ompong occurred during the southwest monsoon season and tracked north of the basin. TY Luis was a category 1 TY that traversed the northern regions of Luzon for 10 h on 14 September 2014. Rainfall brought by TY Luis raised the river level to just above the critical-level threshold at 8.5 m in 61 h. On the other hand, TY Ompong was a very intense category 5 TY that also made landfall over the northern regions of Luzon on 15 September 2018 and stayed on land for 8 h. The rainfall associated with TY Ompong initiated the rise in river elevation until it peaked at a height of 9.01 m. Due to its intensity and large size (~900 km diameter), the associated rainfall also resulted in landslides in the mountainous areas of Cordillera leaving almost 627 222 affected individuals and widespread damage across northern and Central Luzon. The flow of moisture along the prevailing monsoon that passed through the PRB toward TY Luis and TY Ompong extended these TCs’ outer rainbands in the Central Luzon region. The distance of TY Luis and TY Ompong from the PRB means that rainfall from these TCs are less intense. Consequently, the lower rate of increase in water level as seen in the flood onset of their respective hydrographs (Figs. 4b,g) and longer lag times are apparent. The intensity and size of TY Ompong resulted in a faster river response compared to TY Luis, in addition to the fact that the water level was already on the alert level before the onslaught of TY Ompong. The high water level was due to rainfall from another TC, TY Neneng (Barijat), that affected the PRB one week prior to TY Ompong’s landfall. In the cases of TY Ompong and TY Luis, the water level increased from alert to alarm in 6–11 h, more than 2–3 times longer compared to TCs directly crossing the basin. It also took a longer time of 18–37 h from alarm to critical level, but a shorter duration above the critical level from 3 to 12 h.

Other than rainfall intensity, the distribution of precipitation across the basin also plays a role in lag time variations. Flood onset and duration over the basin vary widely from station to station. Figure 6 illustrates the differences in the hydrographs of Arayat and its upstream stations using the case of TY Lando. We calculated the flood wave speed by comparing the occurrences when the water level peaked at different stations. The flood wave took 39 h from the most upstream station, Sapang Buho, to Arayat station and 40 h from Peñaranda station to Arayat station. Both Sapang Buho and Peñaranda stations are located to the northeast of Arayat station. On the other hand, the water level in Zaragoza station located northwest of Arayat station appears to have its peak water level on the same day as Arayat station. This suggests that more rainfall is received in the northeastern subbasins at earlier times which was likely affected by the outer rainbands of TY Lando when it reached its peak intensity off the east coast of Luzon, consistent with the assessment above. Figure 6 also emphasizes the reason for choosing the Arayat station to characterize the hydrologic response of the PRB, its location downstream makes it indicative of flooding in the Central Plain region. Of the eight critical events, flood wave speed from Sapang Buho station to Arayat station ranges from 23 to 50 h.

Fig. 6.
Fig. 6.

Hydrographs of Arayat, Zaragoza, Sapang Buho, Peñaranda, and San Isidro station during the passage of TY Lando (Koppu) (October 2015).

Citation: Journal of Hydrometeorology 22, 4; 10.1175/JHM-D-20-0184.1

3) Remote non-landfalling TC precipitation

Of the eight critical-level events, the combined effects of TS Inday and TD Josie, both non-landfalling TCs, is the only critical event induced by remote TC precipitation. TC-enhanced monsoonal rainfall by TS Inday and TD Josie in Fig. 4h was less intense but persistent over a very long duration of approximately 11 days. This rainfall characteristic resulted in the longest lag time and onset among all critical events leading to a gradually increasing rate of water level that took 49 h to rise from alarm to critical. The effect of continuous rainfall made the river height reach its peak level at 9.04 m.

The speed of catchment response to intense TC rainfall before the river peaks shown in each hydrograph suggests that flood onset and lag times can be in the order of a few hours to a few days. It also implies that rainfall intensity on the basin has different effects on the response time depending on several factors such as TC track, intensity, season, and/or rain field distribution. However, regardless of the TC tracks, accumulated rainfall (AR) for all critical events from alarm to critical level are found to correlate well (r = 0.90, p < 0.05) with the peak water level (WL) given by the linear regression equation WL = 0.012AR + 7.5625. If noncritical events are included, accumulated rainfall from alert to peak level is still significantly correlated (r = 0.75, p < 0.05) with peak water level. This correlation can be used to quantitatively predict river level during TC events. However, prediction of lag time will still be a challenge due to its dependence on other various influencing factors. Generally, TCs that directly traversed the basin had lag times of 1–2 days. Landfalling TCs that did not traverse the basin, moving either north or south of the PRB, had lag times longer than approximately 2–3 days. Last, remote TC rain from non-landfalling TCs has a lag time of approximately 4 days.

4) Other noncritical events

There are other cases of distant TCs that made landfall in the extreme northern Luzon region and caused water level to increase, albeit below the critical level, due to antecedent intense TC rainfall. TY Lawin and TY Mario both did not produce enough heavy rainfall in the PRB but were able to increase the water level to near critical due to increases on top of preceding disturbances. In Fig. 7a, TY Mario increased the water level from 6.6 to 8.4 m 5 days after the onslaught of TY Luis (Fig. 4b). In Fig. 7b, the water level increased from 7.0 to 8.3 m due to TY Lawin 4 days after TY Karen (Fig. 4f). There are on average 1.8 ± 1.2 annual occurrence of consecutive water level peaks reaching at least alert level within a span of one week. This is due to the tendency of consecutive TC cyclogenesis within an equatorial Rossby wave packet found in the WNP basin (Molinari et al. 2007). Consecutive, sequential TCs amplify the effects of even distant TCs with comparatively less intense rainfall.

Fig. 7.
Fig. 7.

Hydrographs of (a) TY Mario (Fung-Wong) and (b) TY Lawin (Haima).

Citation: Journal of Hydrometeorology 22, 4; 10.1175/JHM-D-20-0184.1

We note that while initially high water level prior to a TC onslaught can lead to subsequent flooding, antecedent rainfall (non-TC) itself generally has minimal effect. The calculated API of TY Santi at 116.97 mm is relatively low but has the shortest lag time for all critical events. TY Lando (API = 153.96 mm) and TY Nona (API = 9.85 mm) had similar response times despite a large difference in their API values. Also, TY Ompong (API = 128.61 mm) and TY Luis (API = 150.51 mm) had similar tracks and API values, but shorter response time is observed for the more intense and larger TY Ompong. This suggests that water level response is mainly influenced by the basinwide rain (intensity and amount) regardless of the antecedent watershed soil moisture conditions.

Distant TCs to the south of the basin that made landfall in Visayas can also contribute significant rainfall in the PRB. Figures 8a and 8b show TY Yolanda and TY Marce, respectively, that both made landfall in the central Philippines during the northeast monsoon season. The confluence of moisture flow to their north brought rainfall to the northern flank of these two TCs. This further highlights the importance of TC tracks in relation to the PRB and monsoon season. Hence, the dependence of flooding on TC tracks. Heavy rainfall from consecutive, non-landfalling distant TCs are also able to sustain prolonged alarm level status. Figure 9 shows the hydrographs of TY Maring and TY Nando of 2013. The two TCs sustained water levels just below critical for approximately 126 h. Previous attention is mainly given to TCs that make landfall and directly cross the PRB. From the results of this study, distant TCs with remote precipitation effects can also pose a significant threat to prolonged flooding in the region especially in the recent decade wherein TC-induced remote heavy precipitation events have been shown to be increasing (Bagtasa 2019).

Fig. 8.
Fig. 8.

Hydrographs of (a) TY Yolanda (Haiyan) and (b) TY Marce (Tokage).

Citation: Journal of Hydrometeorology 22, 4; 10.1175/JHM-D-20-0184.1

Fig. 9.
Fig. 9.

Hydrographs of (a) TY Maring (Trami) and (b) TY Nando (Kong-Rey).

Citation: Journal of Hydrometeorology 22, 4; 10.1175/JHM-D-20-0184.1

c. River discharge variability

River discharge is a good indicator for flood risk management and early warning systems. Here, we also calculated the river discharge of the PRB. Figure 10 shows the monthly and annual mean discharge of the river from 2013 to 2018. Figure 10a shows that minimum river discharge is observed during the first 6 months of the year and increases as the rainy season starts. There is also high variability on which month of the year high river discharge occurs. We found that this variability mainly depends on the frequency of TC that affects the PRB. In particular, months with river discharge values above 500 m3 s−1 all have at least two TCs that reached above the alarm level. In August of 2013, precipitation from TY Labuyo and the remote effects of TS Maring and TS Nando caused the high river discharge value for that month. The case is the same for rainfall from TY Luis and TS Mario in September 2014, TS Kabayan and TY Lando in October 2015, STY Lawin and TY Karen in October 2016, TS Henry and TS Inday/TD Josie in July 2018, TS Karding and TS Jolina in August 2018, and TS Neneng and TY Ompong in September 2018. The highest peak and accumulated rainfall for all critical events is from TY Nona (December 2015). It is interesting to note that even without directly passing through the PRB, TY Nona’s intense wind and extended rainbands to its north brought such intense rainfall that the river discharge in that month peaked at 480 m3 s−1. The number of TCs that resulted in at least an alarm level event in a year is found to significantly correlate (r = 0.98, p < 0.05) with annual mean river discharge. This emphasizes the significant contribution of TC-induced rain in river flooding of the PRB. River discharge of the PRB shown in Fig. 10b is observed to have an increasing trend of 7% (p < 0.05) in the 6-yr observation. However, the study period is not enough to provide long-term, conclusive trend information. The highest annual mean discharge within the study period was observed in 2018 while the lowest is in 2017.

Fig. 10.
Fig. 10.

(a) Monthly and (b) annual mean discharge at Arayat station.

Citation: Journal of Hydrometeorology 22, 4; 10.1175/JHM-D-20-0184.1

d. Checklist for qualitative flood forecasting

We propose a simple checklist flood forecast method for the PRB based on the events of the study period. Table 5 shows the possible resulting peak river level depending on TC track location, seasons, and basinwide accumulated rainfall. TCs directly crossing the PRB with rainfall greater than 90 mm yield critical-level events. Intense TCs making landfall elsewhere in the Philippines can lead to any level depending on the accumulated rainfall when north-tracking TCs occur during the southwest and south-tracking TCs occur during the northeast monsoon. In addition, north-tracking TCs can reach alarm levels in the northeast monsoon season if the TC is intense and/or has a large radius. For non-landfalling TCs, flooding events can occur at any level during the southwest monsoon season, however, higher accumulated rainfall is needed to raise water levels due to the comparatively lower rain rate of remote TC precipitation.

Table 5.

Checklist for PRB flood level forecasting involving TC track, season, and basinwide accumulated rainfall. An asterisk indicates an intense/large-radius TC.

Table 5.

4. Summary and conclusions

Basin hydrological response such as the rate of onset of flooding and response of a catchment to intense rainfall is important for flood forecasting. With extensive knowledge of a basin’s response, the affected population can be warned ahead of an impending hazard that a certain storm might bring. This study focused on the hydrological response of the Pampanga River basin, located in Central Luzon in the Philippines, to heavy precipitation induced by tropical cyclones. Flooding is characterized by river water height at Arayat station with alert (5 m), alarm (6 m), and critical (8.5 m) levels, when the river is 40%, 60%, and 100% full, respectively. In the analysis of the 6-yr gauge data from 2013 to 2018, all flooding events in the PRB at different intensity levels were caused by TC-induced precipitation. All intense TCs that directly traversed the basin resulted in critical-level flooding. The tracks of TCs that made landfall elsewhere in the Philippines and induced flooding in the basin are dependent on the season. Flooding is induced when TCs move to the north of the PRB during the southwest monsoon and to the south of the PRB during the northeast monsoon season. This is due to the TC feeder bands forming along the confluence of monsoonal flow and TC circulation located over the PRB. Moreover, remote non-landfalling TCs which are known to produce heavy precipitation events on the island of Luzon during the southwest monsoon season can also induce flooding, with a higher chance (>50%) of flooding at the alert and alarm levels.

TCs directly traversing the basin had lag times of 1–2 days and the shortest flood onset of 7–27 h from alert to critical. TCs that made landfall elsewhere in the Philippines had longer lag times of approximately 2–3 days and onset from alert to critical of 11–48 h, with shorter lag and onset times for very intense TCs. Last, remote TC rain from non-landfalling TCs that induce flooding has a lag time of approximately 4 days due to its persistent but lower rain rate. The differences in lag times are partly influenced by TC intensity where more intense TCs possessing higher rain rates (Lin et al. 2015) result in shorter lag times. The flood onset and lag times are based only on the 6-yr rain and river level dataset analyzed in this study. Longer-term analysis can further improve this quantitative assessment of river response and can help resolve the current ambiguity in lag time determination, particularly for non-landfalling TCs. A longer study period can also provide more information on the influence of TC track variations (Santos et al. 2018). Another study using long-term flood modeling data, which has been shown to be robust enough in filling data gaps after certain bias corrections under Philippine climate regimes (Ibarra et al. 2020), is underway.

In the present study, we also proposed a simple checklist flood forecasting method based on the TCs that affected the PRB from 2013 to 2018. Flood level is qualitatively derived from TC track position in relation to the PRB, season, and basinwide accumulated rainfall, which can be further refined and improved as more events populate the PRB-associated hydrometeorological data in the future.

Acknowledgments

The authors would like to acknowledge the DOST-SEI-ASTHRDP scholarship for supporting the first author’s Ph.D. program. We also thank Mr. Hilton Hernando (Asst. weather services chief) of the Pampanga River Basin Flood Forecasting and Warning Center for providing the data, Dr. Guillermo Q. Tabios III for his useful comments, and the anonymous reviewers for their constructive comments.

REFERENCES

  • Akasaka, I., W. Morishima, and T. Mikami, 2007: Seasonal march and its spatial difference of rainfall in the Philippines. Int. J. Climatol., 27, 715725, https://doi.org/10.1002/joc.1428.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Archer, D., and H. Fowler, 2018: Characterising flash flood response to intense rainfall and impacts using historical information and gauged data in Britain. J. Flood Risk Manage., 11, S121S133, https://doi.org/10.1111/jfr3.12187.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bagtasa, G., 2017: Contribution of tropical cyclones to rainfall in the Philippines. J. Climate, 30, 36213633, https://doi.org/10.1175/JCLI-D-16-0150.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bagtasa, G., 2019: Enhancement of summer monsoon rainfall by tropical cyclones in northwestern Philippines. J. Meteor. Soc. Japan, 97, 967976, https://doi.org/10.2151/jmsj.2019-052.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bagtasa, G., 2020a: 118-year climate and extreme weather events of Metropolitan Manila in the Philippines. Int. J. Climatol., 40, 12281240, https://doi.org/10.1002/joc.6267.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bagtasa, G., 2020b: Influence of Madden–Julian oscillation on the intraseasonal variability of summer and winter monsoon rainfall in the Philippines. J. Climate, 33, 95819594, https://doi.org/10.1175/JCLI-D-20-0305.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Cayanan, E. O., T.-C. Chen, J. C. Argete, M.-C. Yen, and P. D. Nilo, 2011: The effect of tropical cyclones on southwest monsoon rainfall in the Philippines. J. Meteor. Soc. Japan, 89, 123139, https://doi.org/10.2151/jmsj.2011-A08.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • DENR, 2019: Regional profile: Region 3. Department of Environment and Natural Resources, https://r3.denr.gov.ph/index.php/about-us/regional-profile.

  • Emanuel, K., 2005: Increasing destructiveness of tropical cyclones over the past 30 years. Nature, 436, 686688, https://doi.org/10.1038/nature03906.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ferraris, C. C., 1971: An Analysis of Rainfall and Floods in the Pampanga River Basin (1946-1966). Tech. Rep., Philippine Weather Bureau, 125 pp.

  • Fiedler, F. R., 2003: Simple, practical method for determining station weights using Thiessen polygons and isohyetal maps. J. Hydrol. Eng., 8, 219221, https://doi.org/10.1061/(ASCE)1084-0699(2003)8:4(219).

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gebrehiwot, S. G., U. Ilstedt, A. Gardenas, and K. Bishop, 2011: Hydrological characterization of watersheds in the Blue Nile Basin, Ethiopia. Hydrol. Earth Syst. Sci., 15, 1120, https://doi.org/10.5194/hessd-7-4089-2010.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gordon, A., 1973: The great Philippine floods of 1972. Weather, 28, 404415, https://doi.org/10.1002/j.1477-8696.1973.tb00793.x.

  • Ibarra, D. E., C. P. C. David, and P. L. M. Tolentino, 2020: Evaluation and bias correction of an observations-based global runoff dataset using historical streamflow observations from small tropical catchments in the Philippines. Hydrol. Earth Syst. Sci. Discuss., https://doi.org/10.5194/hess-2020-26.

    • Search Google Scholar
    • Export Citation
  • Jaranilla-Sanchez, P. A., T. Koike, C. T. Nyunt, M. Rasmy, I. Hasegawa, A. Matsumura, and D. Ogawada, 2013: Hydrological impacts of a changing climate on floods and droughts in Philippine river basins. J. Japan Soc. Civil Eng., 69B, I_13I_18, https://doi.org/10.2208/jscejhe.69.I_13.

    • Search Google Scholar
    • Export Citation
  • Jiang, H., E. M. Ramirez, and D. J. Cecil, 2013: Convective and rainfall properties of tropical cyclone inner cores and rainbands from 11 years of TRMM data. Mon. Wea. Rev., 141, 431450, https://doi.org/10.1175/MWR-D-11-00360.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • JICA, 2011: The study on integrated water resources management for poverty alleviation and economic development in the Pampanga River Basin in the Republic of the Philippines. Japan International Cooperation Agency, 350 pp., https://openjicareport.jica.go.jp/pdf/12018974_01.pdf.

  • Kudo, T., R. Kawamura, H. Hirata, K. Ichiyanagi, M. Tanoue, and K. Yoshimura, 2014: Large-scale vapor transport of remotely evaporated seawater by a Rossby wave response to typhoon forcing during the Baiu/Meiyu season as revealed by the JRA-55 reanalysis. J. Geophys. Res. Atmos., 119, 88258838, https://doi.org/10.1002/2014JD021999.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lee, D., B. Oh, H. Kim, S. Lee, and G. Chung, 2013: Comparison of the hydro-climatological characteristics for the extra-ordinary flood induced by tropical cyclone in the selected river basins. Trop. Cyclone Res. Rev., 2, 4554, https://doi.org/10.6057/2013TCRR01.05.

    • Search Google Scholar
    • Export Citation
  • Lin, Y., M. Zhao, and M. Zhang, 2015: Tropical cyclone rainfall area controlled by relative sea surface temperature. Nat. Commun., 6, 6591, https://doi.org/10.1038/ncomms7591.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Linsley, R. K., Jr., M. A. Kohler, and J. L. Paulhus, 1975: Hydrology for Engineers. McGraw-Hill, 508 pp.

  • McDonnell, J., and Coauthors, 2007: Moving beyond heterogeneity and process complexity: A new vision for watershed hydrology. Water Resour. Res., 43, W07301, https://doi.org/10.1029/2006WR005467.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Molinari, J., K. Lombardo, and D. Vollaro, 2007: Tropical cyclogenesis within an equatorial Rossby wave packet. J. Atmos. Sci., 64, 13011317, https://doi.org/10.1175/JAS3902.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Moron, V., A. Lucero, F. Hilario, B. Lyon, A. Robertson, and D. DeWitt, 2009: Spatio-temporal variability and predictability of summer monsoon onset over the Philippines. Climate Dyn., 33, 11591177, https://doi.org/10.1007/s00382-008-0520-5.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • NDRRMC, 2011: SitRep No. 17 re Effects of Typhoon “PEDRING” (Nesat). National Disaster Risk Reduction and Management Council, 34 pp., https://ndrrmc.gov.ph/attachments/article/1748/SitRep_No_17_re_Effects_of_Typhoon_PEDRING_as_of_03OCT2011_0600H.pdf.

  • NDRRMC, 2013: SitRep No. 10 re Effects of Typhoon “Santi” (Nari). National Disaster Risk Reduction and Management Council, 31 pp., https://reliefweb.int/report/philippines/ndrrmc-update-sitrep-no-10-re-effects-typhoon-santi-nari.

  • NDRRMC, 2015: Final Report re Preparedness measures and Effects of Typhoon “LANDO” (Koppu). National Disaster Risk Reduction and Management Council, 136 pp., http://ndrrmc.gov.ph/attachments/article/2607/FINAL_REPORT_re_Preparedness_Measures_and_Effects_of_Typhoon_LANDO_KOPPU_as_of_14_-_21OCT2015.pdf.

  • NEDA, 2018: Central Luzon’s GRDP grows 9.3 percent, sustains high growth in 2017. National Economic Development Agency, http://nro3.neda.gov.ph/central-luzons-grdp-grows-9-3-percent-sustains-high-growth-in-2017/.

  • Okazumi, T., M. Miyamoto, B. B. Shrestha, and M. Gusyev, 2014: Uncertainty estimation during the process of flood risk assessment in developing countries–Case study in the Pampanga River basin. J. Disaster Res., 9, 6977, https://doi.org/10.20965/jdr.2014.p0069.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Penna, D., H. Van Meerveld, G. Zuecco, G. Dalla Fontana, and M. Borga, 2016: Hydrological response of an Alpine catchment to rainfall and snowmelt events. J. Hydrol., 537, 382397, https://doi.org/10.1016/j.jhydrol.2016.03.040.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • PRBFFWC, 2015: Post-Flood Report 2015. Pampanga River Basin Flood Forecasting and Warning Center, http://prffwc.synthasite.com/resources/annual%20report%202015-PRB.pdf.

  • PSA, 2011: Central Luzon’s population hits 11.2M. Philippine Statistics Authority, http://rsso03.psa.gov.ph/article/central-luzon’s-population-hits-112m-result-2015-census-population.

  • Santos, M. S., L. Mediero, C. H. R. Lima, and L. Z. Moura, 2018: Links between different classes of storm tracks and the flood trends in Spain. J. Hydrol., 567, 7185, https://doi.org/10.1016/j.jhydrol.2018.10.003.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Schopka, H., L. Derry, and C. Arcilla, 2011: Chemical weathering, river geochemistry and atmospheric carbon fluxes from volcanic and ultramafic regions on Luzon Island, the Philippines. Geochim. Cosmochim. Acta, 75, 9781002, https://doi.org/10.1016/j.gca.2010.11.014.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Shrestha, B. B., T. Okazumi, M. Miyamoto, and H. Sawano, 2016: Flood damage assessment in the Pampanga River basin of the Philippines. J. Flood Risk Manage., 9, 355369, https://doi.org/10.1111/jfr3.12174.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Tolentino, P. L. M., A. Poortinga, H. Kanamaru, S. Keesstra, J. Maroulis, C. P. C. David, and C. J. Ritsema, 2016: Projected impact of climate change on hydrological regimes in the Philippines. PLOS One, 11, e0163941, https://doi.org/10.1371/journal.pone.0163941

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Villafuerte, M. Q., I. Macadam, J. Daron, J. Katzfey, T. A. Cinco, E. D. Ares, and R. G. Jones, 2020: Projected changes in rainfall and temperature over the Philippines from multiple dynamical downscaling models. Int. J. Climatol., 40, 17841804, https://doi.org/10.1002/joc.6301.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wang, B., 2002: Rainy season of the Asian–Pacific summer monsoon. J. Climate, 15, 386398, https://doi.org/10.1175/1520-0442(2002)015<0386:RSOTAP>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Webster, P. J., G. J. Holland, J. A. Curry, and H.-R. Chang, 2005: Changes in tropical cyclone number, duration, and intensity in a warming environment. Science, 309, 18441846, https://doi.org/10.1126/science.1116448.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Yumul, G. P., C. B. Dimalanta, N. T. Servando, and N. A. Cruz, 2013: Abnormal weather events in 2009, increased precipitation and disastrous impacts in the Philippines. Climatic Change, 118, 715727, https://doi.org/10.1007/s10584-012-0661-8.

    • Crossref
    • Search Google Scholar
    • Export Citation
Save
  • Akasaka, I., W. Morishima, and T. Mikami, 2007: Seasonal march and its spatial difference of rainfall in the Philippines. Int. J. Climatol., 27, 715725, https://doi.org/10.1002/joc.1428.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Archer, D., and H. Fowler, 2018: Characterising flash flood response to intense rainfall and impacts using historical information and gauged data in Britain. J. Flood Risk Manage., 11, S121S133, https://doi.org/10.1111/jfr3.12187.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bagtasa, G., 2017: Contribution of tropical cyclones to rainfall in the Philippines. J. Climate, 30, 36213633, https://doi.org/10.1175/JCLI-D-16-0150.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bagtasa, G., 2019: Enhancement of summer monsoon rainfall by tropical cyclones in northwestern Philippines. J. Meteor. Soc. Japan, 97, 967976, https://doi.org/10.2151/jmsj.2019-052.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bagtasa, G., 2020a: 118-year climate and extreme weather events of Metropolitan Manila in the Philippines. Int. J. Climatol., 40, 12281240, https://doi.org/10.1002/joc.6267.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bagtasa, G., 2020b: Influence of Madden–Julian oscillation on the intraseasonal variability of summer and winter monsoon rainfall in the Philippines. J. Climate, 33, 95819594, https://doi.org/10.1175/JCLI-D-20-0305.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Cayanan, E. O., T.-C. Chen, J. C. Argete, M.-C. Yen, and P. D. Nilo, 2011: The effect of tropical cyclones on southwest monsoon rainfall in the Philippines. J. Meteor. Soc. Japan, 89, 123139, https://doi.org/10.2151/jmsj.2011-A08.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • DENR, 2019: Regional profile: Region 3. Department of Environment and Natural Resources, https://r3.denr.gov.ph/index.php/about-us/regional-profile.

  • Emanuel, K., 2005: Increasing destructiveness of tropical cyclones over the past 30 years. Nature, 436, 686688, https://doi.org/10.1038/nature03906.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ferraris, C. C., 1971: An Analysis of Rainfall and Floods in the Pampanga River Basin (1946-1966). Tech. Rep., Philippine Weather Bureau, 125 pp.

  • Fiedler, F. R., 2003: Simple, practical method for determining station weights using Thiessen polygons and isohyetal maps. J. Hydrol. Eng., 8, 219221, https://doi.org/10.1061/(ASCE)1084-0699(2003)8:4(219).

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gebrehiwot, S. G., U. Ilstedt, A. Gardenas, and K. Bishop, 2011: Hydrological characterization of watersheds in the Blue Nile Basin, Ethiopia. Hydrol. Earth Syst. Sci., 15, 1120, https://doi.org/10.5194/hessd-7-4089-2010.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gordon, A., 1973: The great Philippine floods of 1972. Weather, 28, 404415, https://doi.org/10.1002/j.1477-8696.1973.tb00793.x.

  • Ibarra, D. E., C. P. C. David, and P. L. M. Tolentino, 2020: Evaluation and bias correction of an observations-based global runoff dataset using historical streamflow observations from small tropical catchments in the Philippines. Hydrol. Earth Syst. Sci. Discuss., https://doi.org/10.5194/hess-2020-26.

    • Search Google Scholar
    • Export Citation
  • Jaranilla-Sanchez, P. A., T. Koike, C. T. Nyunt, M. Rasmy, I. Hasegawa, A. Matsumura, and D. Ogawada, 2013: Hydrological impacts of a changing climate on floods and droughts in Philippine river basins. J. Japan Soc. Civil Eng., 69B, I_13I_18, https://doi.org/10.2208/jscejhe.69.I_13.

    • Search Google Scholar
    • Export Citation
  • Jiang, H., E. M. Ramirez, and D. J. Cecil, 2013: Convective and rainfall properties of tropical cyclone inner cores and rainbands from 11 years of TRMM data. Mon. Wea. Rev., 141, 431450, https://doi.org/10.1175/MWR-D-11-00360.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • JICA, 2011: The study on integrated water resources management for poverty alleviation and economic development in the Pampanga River Basin in the Republic of the Philippines. Japan International Cooperation Agency, 350 pp., https://openjicareport.jica.go.jp/pdf/12018974_01.pdf.

  • Kudo, T., R. Kawamura, H. Hirata, K. Ichiyanagi, M. Tanoue, and K. Yoshimura, 2014: Large-scale vapor transport of remotely evaporated seawater by a Rossby wave response to typhoon forcing during the Baiu/Meiyu season as revealed by the JRA-55 reanalysis. J. Geophys. Res. Atmos., 119, 88258838, https://doi.org/10.1002/2014JD021999.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lee, D., B. Oh, H. Kim, S. Lee, and G. Chung, 2013: Comparison of the hydro-climatological characteristics for the extra-ordinary flood induced by tropical cyclone in the selected river basins. Trop. Cyclone Res. Rev., 2, 4554, https://doi.org/10.6057/2013TCRR01.05.

    • Search Google Scholar
    • Export Citation
  • Lin, Y., M. Zhao, and M. Zhang, 2015: Tropical cyclone rainfall area controlled by relative sea surface temperature. Nat. Commun., 6, 6591, https://doi.org/10.1038/ncomms7591.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Linsley, R. K., Jr., M. A. Kohler, and J. L. Paulhus, 1975: Hydrology for Engineers. McGraw-Hill, 508 pp.

  • McDonnell, J., and Coauthors, 2007: Moving beyond heterogeneity and process complexity: A new vision for watershed hydrology. Water Resour. Res., 43, W07301, https://doi.org/10.1029/2006WR005467.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Molinari, J., K. Lombardo, and D. Vollaro, 2007: Tropical cyclogenesis within an equatorial Rossby wave packet. J. Atmos. Sci., 64, 13011317, https://doi.org/10.1175/JAS3902.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Moron, V., A. Lucero, F. Hilario, B. Lyon, A. Robertson, and D. DeWitt, 2009: Spatio-temporal variability and predictability of summer monsoon onset over the Philippines. Climate Dyn., 33, 11591177, https://doi.org/10.1007/s00382-008-0520-5.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • NDRRMC, 2011: SitRep No. 17 re Effects of Typhoon “PEDRING” (Nesat). National Disaster Risk Reduction and Management Council, 34 pp., https://ndrrmc.gov.ph/attachments/article/1748/SitRep_No_17_re_Effects_of_Typhoon_PEDRING_as_of_03OCT2011_0600H.pdf.

  • NDRRMC, 2013: SitRep No. 10 re Effects of Typhoon “Santi” (Nari). National Disaster Risk Reduction and Management Council, 31 pp., https://reliefweb.int/report/philippines/ndrrmc-update-sitrep-no-10-re-effects-typhoon-santi-nari.

  • NDRRMC, 2015: Final Report re Preparedness measures and Effects of Typhoon “LANDO” (Koppu). National Disaster Risk Reduction and Management Council, 136 pp., http://ndrrmc.gov.ph/attachments/article/2607/FINAL_REPORT_re_Preparedness_Measures_and_Effects_of_Typhoon_LANDO_KOPPU_as_of_14_-_21OCT2015.pdf.

  • NEDA, 2018: Central Luzon’s GRDP grows 9.3 percent, sustains high growth in 2017. National Economic Development Agency, http://nro3.neda.gov.ph/central-luzons-grdp-grows-9-3-percent-sustains-high-growth-in-2017/.

  • Okazumi, T., M. Miyamoto, B. B. Shrestha, and M. Gusyev, 2014: Uncertainty estimation during the process of flood risk assessment in developing countries–Case study in the Pampanga River basin. J. Disaster Res., 9, 6977, https://doi.org/10.20965/jdr.2014.p0069.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Penna, D., H. Van Meerveld, G. Zuecco, G. Dalla Fontana, and M. Borga, 2016: Hydrological response of an Alpine catchment to rainfall and snowmelt events. J. Hydrol., 537, 382397, https://doi.org/10.1016/j.jhydrol.2016.03.040.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • PRBFFWC, 2015: Post-Flood Report 2015. Pampanga River Basin Flood Forecasting and Warning Center, http://prffwc.synthasite.com/resources/annual%20report%202015-PRB.pdf.

  • PSA, 2011: Central Luzon’s population hits 11.2M. Philippine Statistics Authority, http://rsso03.psa.gov.ph/article/central-luzon’s-population-hits-112m-result-2015-census-population.

  • Santos, M. S., L. Mediero, C. H. R. Lima, and L. Z. Moura, 2018: Links between different classes of storm tracks and the flood trends in Spain. J. Hydrol., 567, 7185, https://doi.org/10.1016/j.jhydrol.2018.10.003.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Schopka, H., L. Derry, and C. Arcilla, 2011: Chemical weathering, river geochemistry and atmospheric carbon fluxes from volcanic and ultramafic regions on Luzon Island, the Philippines. Geochim. Cosmochim. Acta, 75, 9781002, https://doi.org/10.1016/j.gca.2010.11.014.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Shrestha, B. B., T. Okazumi, M. Miyamoto, and H. Sawano, 2016: Flood damage assessment in the Pampanga River basin of the Philippines. J. Flood Risk Manage., 9, 355369, https://doi.org/10.1111/jfr3.12174.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Tolentino, P. L. M., A. Poortinga, H. Kanamaru, S. Keesstra, J. Maroulis, C. P. C. David, and C. J. Ritsema, 2016: Projected impact of climate change on hydrological regimes in the Philippines. PLOS One, 11, e0163941, https://doi.org/10.1371/journal.pone.0163941

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Villafuerte, M. Q., I. Macadam, J. Daron, J. Katzfey, T. A. Cinco, E. D. Ares, and R. G. Jones, 2020: Projected changes in rainfall and temperature over the Philippines from multiple dynamical downscaling models. Int. J. Climatol., 40, 17841804, https://doi.org/10.1002/joc.6301.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wang, B., 2002: Rainy season of the Asian–Pacific summer monsoon. J. Climate, 15, 386398, https://doi.org/10.1175/1520-0442(2002)015<0386:RSOTAP>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Webster, P. J., G. J. Holland, J. A. Curry, and H.-R. Chang, 2005: Changes in tropical cyclone number, duration, and intensity in a warming environment. Science, 309, 18441846, https://doi.org/10.1126/science.1116448.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Yumul, G. P., C. B. Dimalanta, N. T. Servando, and N. A. Cruz, 2013: Abnormal weather events in 2009, increased precipitation and disastrous impacts in the Philippines. Climatic Change, 118, 715727, https://doi.org/10.1007/s10584-012-0661-8.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Fig. 1.

    PRB topography with provincial boundaries and location of PAGASA gauging stations. Blue triangles are rainfall stations while the red triangles are combined rainfall and water level stations.

  • Fig. 2.

    Hydrograph data from 2013 to 2018 in PRB showing identified rain/flood events based on exceedance of alert (yellow line), alarm (orange), and critical level (red line). Black bars denote basin mean rainfall and dark blue lines depict water level while pink boxes highlight critical-level events with tracks of coincident tropical cyclones that entered the Philippine Area of Responsibility. Tracks are from the TCGS of PAGASA.

  • Fig. 3.

    Tracks, local (and international) names, classification, and dates of the TCs that caused alert and alarm levels in PRB. Tracks are from the TCGS of PAGASA.

  • Fig. 4.

    Hydrographs during the passage of TCs (a) TY Santi (Nari), (b) TY Luis (Kalmaegi), (c) TS Kabayan (Mujigae), (d) TY Lando (Koppu), (e) TY Nona (Melor), (f) TY Karen (Sarika), (g) the combined effect of TS Inday (Ampil) and TD Josie TY, and (h) Ompong (Mangkhut).

  • Fig. 5.

    Time interval between water levels from alert to alarm (yellow), from alarm to critical (orange), and from critical to peak level (red) in hours.

  • Fig. 6.

    Hydrographs of Arayat, Zaragoza, Sapang Buho, Peñaranda, and San Isidro station during the passage of TY Lando (Koppu) (October 2015).

  • Fig. 7.

    Hydrographs of (a) TY Mario (Fung-Wong) and (b) TY Lawin (Haima).

  • Fig. 8.

    Hydrographs of (a) TY Yolanda (Haiyan) and (b) TY Marce (Tokage).

  • Fig. 9.

    Hydrographs of (a) TY Maring (Trami) and (b) TY Nando (Kong-Rey).

  • Fig. 10.

    (a) Monthly and (b) annual mean discharge at Arayat station.

All Time Past Year Past 30 Days
Abstract Views 1859 0 0
Full Text Views 4186 2163 172
PDF Downloads 4074 1995 158