1. Introduction
Changes in global rainfall patterns are of great concern in the midst of our warming climate. Studies of rainfall patterns in global and regional scales have shown variations in annual precipitation trends for different regions (Chadwick et al. 2013; Liu et al. 2013; Zhang and Zhou 2011; Wang et al. 2012). In the Philippines, where the agricultural sector employs a third of its workforce (World Bank 2015), rain is the most important daily weather phenomenon. The Philippines is an archipelago consisting of more than 7100 islands situated between the western rim of the Pacific Ocean and the South China Sea (locally known as West Philippine Sea). It is dominated by complex terrain, with long mountain ranges on the two largest islands of Luzon and Mindanao to the north and south, respectively, and Visayas consists of smaller islands in the central Philippines. Its location subjects it to numerous tropical cyclones (TC) that form in the northwestern Pacific (NWP) basin (Weinkle et al. 2012). An average of 19.4 TCs enter the Philippine Area of Responsibility bounded by coordinates 25°N–120°E, 25°N–135°E, 5°N–135°E, 5°N–115°E, 15°N–115°E, and 21°N–120°E (dashed line in Fig. 1). Around half of these make landfall each year (Cinco et al. 2016).
The Philippine Area of Responsibility (enclosed by the dashed line) and TC count density in the northwestern Pacific for (a) MAM, (b) JJA, (c) SON, and (d) DJF. The size of the gray rectangles extends from a count of 1 to 14 in increments of 1.
Citation: Journal of Climate 30, 10; 10.1175/JCLI-D-16-0150.1
Recently, several studies looked into the local trends and variabilities of temperature and rainfall to elucidate climatic observations that can be vital for climate change adaptation measures. Cruz et al. (2013) showed a decreasing rainfall trend in the western portion of the Philippines during the boreal summer or southwest monsoon season from 1960 to 2010. Their results also indicate a prolonged dry season for most of the western Philippines. Villafuerte et al. (2014) analyzed long-term trends and the variability of rainfall extremes using seven extreme precipitation indices (EPI). They found a drying trend from January to March and an increasing trend during extreme rainfall events from July to September. Longer trend analysis from 1911 to 2010 by Villafuerte et al. (2014) indicates the southwestward extension of the northwestern Pacific subtropical high (NWPSH) and the weakening of westerly winds as causes of this drying trend. Cinco et al. (2014) found mostly warming trends in surface temperature, as well as an increase in extreme temperature and daily rainfall events, from 37 synoptic stations. Significant increases were distributed all over the country. TC contribution on rainfall, however, was not considered in these studies.
The climate of the Philippines is highly influenced by ENSO (Hilario et al. 2009; Ropelewski and Halpert 1996). Drier (wetter) conditions are associated with El Niño (La Niña) events. Lyon and Camargo (2009) observed a seasonal reversal in ENSO rainfall signal in the Philippines. Above- (below) average rainfall is seen in El Niño (La Niña) years during boreal summer, and a reverse rainfall anomaly is seen the following fall. TC activity is also shown to be enhanced (reduced) during boreal summer of El Niño (La Niña) events due to changes in midlevel atmospheric moisture. A number of papers also reported on the relationship of TCs and ENSO in the NWP basin (Atkinson 1977; Camargo and Sobel 2005; Chan 2000; Corporal-Lodangco and Leslie 2016; Wang and Chan 2002). In general, TC geneses tend to occur farther to the southeast during El Niño years (Corporal-Lodangco and Leslie 2016; Wang and Chan 2002). As a result, TCs have longer lifetimes over the Pacific. This leads to more intensification and a tendency to track toward the northwest quadrant of the NWP. Consequently, a northwesterly track leads to less TC directly affecting the Philippines (Hilario et al. 2009).
Tropical cyclones are regarded as the most destructive hydrometeorological hazard in the Philippines. TCs in the last four decades have an estimated normalized cost of more than $2 billion in damages (Cinco et al. 2016) and thousands of lost lives. Historic accounts of TCs starting from the sixteenth century emphasize the destructive nature of this hydrometeorological event (Ribera et al. 2008). Aside from their destructive potential, TCs can also have a significant contribution to the terrestrial hydrologic cycle (Coronas 1912; Flores and Balagot 1969). Moreover, understanding its impacts is vital to the proper management of water resources (Dare et al. 2012; Ren et al. 2006).
Despite the importance of TCs in the Philippines, there is limited literature on local TC-induced rainfall (TC rain) contribution. Kubota and Wang (2009) studied the effects of TCs on rainfall in the NWP region using 22 stations, including 7 stations from the Philippines. They showed that the interannual variability of rainfall in the NWP is modulated by ENSO. Their results suggest that rainfall variability is controlled by changes in nonlocal circulations that modify TC genesis location and tracks. Cayanan et al. (2011) reported cases of an indirect effect of TCs located to the northeast of Luzon, northern Philippines. Heavy rainfall events were caused by the interaction of TCs and southwesterly wind during the boreal summer monsoon period. Enhanced moisture flow coupled with local orography generated active convection along the western, windward side of Luzon. Recently, Cinco et al. (2016) made an initial analysis of local TC-induced rain. The largest contribution of up to 50% was observed in the northern Philippines. However, their analyses focused on TC landfall frequency; no trends were found.
The aim of this work is to quantify and characterize the amount of rainfall contributed by TCs using long-term gridded precipitation datasets over different seasons and to investigate the variability and trends in regions with distinct rainfall climate regimes. The next section will discuss the precipitation datasets used, TC track information, and the method used to compute TC-induced rainfall. In section 3, the distribution of TC-contributed rainfall, its trends, and variabilities are presented. The influence of TCs on southwest monsoon rainfall is also considered. Section 4 summarizes the result of this study.
2. Data and method
This study analyzed TC-induced precipitation from 1951 to 2014. Among the precipitation data used is the Asian Precipitation–Highly Resolved Observational Data Integration Toward Evaluation of the Water Resources (APHRODITE) Monsoon Asia, version 1101R2 (Yatagai et al. 2012). This dataset is composed of daily gridded precipitation covering Asia, including Southeast Asia, with a grid resolution of 0.25° × 0.25°. The data cover a period of 57 yr from 1951 to 2007. Rainfall from this dataset is derived from rain gauges that are up to 4.5 times denser than data available through the Global Telecommunication System (GTS) of the World Weather Watch (Yasutomi et al. 2011).
To cover the years 2008–14, daily precipitation data from TRMM 3B42, version 7, for the period 1998–2014 were used. The dataset also uses a grid resolution of 0.25° × 0.25°. Details are found on its website (http://mirador.gsfc.nasa.gov/collections/TRMM_3B42_daily__007.shtml). The TRMM Multisatellite Precipitation Analysis is widely used in climatological rainfall studies due to its long temporal and global spatial coverage (Liu 2015; Breña-Naranjo et al. 2015; Wang et al. 2014).
To address concerns over using two different datasets, both APHRODITE and TRMM data used the same grids in the analysis. A comparison of annual rainfall of the two datasets over a 10-yr overlap from 1998 to 2007 shows a statistically significant correlation of r = 0.96 at the 95% significance level (p < 0.05). However, the TRMM rainfall values have a mean wet bias of 12.1% compared to APHRODITE. Bias offset correction was applied to the TRMM data for consistency. Figure 2 shows the comparison between observed mean annual rainfall with the bias-corrected TRMM and APHRODITE datasets per climate subtype. The scatterplot in Fig. 2 reveals that both datasets behave similarly and correlate highly (r = 0.85) with observation. Observed precipitation data provided by the National Climatic Data Center’s Global Summary of the Day (http://www1.ncdc.noaa.gov/pub/data/gsod) from 1990 to 2014 was used to assess the two datasets. A total of 31 synoptic stations distributed across the Philippines were included in the assessment.
Observed vs gridded precipitation data (TRMM: red, APHRODITE: black).
Citation: Journal of Climate 30, 10; 10.1175/JCLI-D-16-0150.1
TC reanalyzed best track data are from the website of the Regional Specialized Meteorological Centre (RSMC) Tokyo-Typhoon Center of the Japan Meteorological Agency (http://www.jma.go.jp/jma/jma-eng/jma-center/rsmc-hp-pub-eg/trackarchives.html). A total of 1673 tropical cyclones over the same period as the precipitation datasets (1951–2014) in the NWP basin were used in the analysis. Wind, geopotential height, and total column water vapor flux data at 1.25° × 1.25° resolution are from the Japanese 55-year Reanalysis (JRA-55) project, carried out by the Japan Meteorological Agency (JMA 2013).
Rainfall within a 10° (~1100 km) radius from the TC center was chosen in this study to be rainfall associated with TCs. Mean total rainfall within varying distance radii from the TC center was calculated for the whole study period. Similar to the results of Jiang et al. (2008) and Kubota and Wang (2009), Fig. 3 shows the mean rainfall amount decreases with a larger TC influence radius and becomes almost constant from around a 10° radius onward. Hence, this was selected as the optimum range of TC-induced rainfall.
Mean total rainfall vs radius from the TC center.
Citation: Journal of Climate 30, 10; 10.1175/JCLI-D-16-0150.1
3. Results and discussion
a. TC–induced rainfall
The distribution of total averaged daily rainfall, TC rain, and TC percentage contribution from 1951 to 2014 in the Philippines is shown in Fig. 4. The TC percentage contribution is computed as the ratio of TC rain to total rainfall multiplied by 100%. The eastern Philippines region is seen in Fig. 4a to have the most amount of rain at an annual average of 3700 mm or 10 mm day−1. Heavy rain in this region is due to orographic land–air interaction during the boreal winter monsoon period (Flores and Balagot 1969). TC-induced rainfall (Fig. 4b) for all years is apparent in the northern half of the country and most pronounced along western Luzon. Consequently, the TC percentage contribution (Fig. 4c) is more than 40% in most of Luzon, and highest at 54.2% along the coastal regions of the northwest. On the other hand, the northeast region of Mindanao, which experiences heavy annual rainfall of about 3600 mm on average, receives only 13.5% rain contribution from TCs. The lowest contribution is seen in the south to southwest Mindanao at 6.2%.
Mean total column water vapor flux (arrows and gray shades) for (a) JJA, (b) cases where a TC is within 19°–21°N, 125°–127°E, and (c) difference (b) − (a). The gray scale is white for a value of 0 and goes to black for a value of 1000 in increments of 100 kg m−1 s−1.
Citation: Journal of Climate 30, 10; 10.1175/JCLI-D-16-0150.1
TC occurrence in the NWP basin varies seasonally. The quarterly averaged number of TCs for December–February (DJF) of the following year is 1.8 TCs, MAM at 2.25 TCs, JJA at 11.28 TCs, and SON at 10.8 TCs. Figure 1 shows the corresponding spatial TC count density distribution in the NWP basin. High TC activity is seen in both JJA and SON. However, the months of SON show a more southwesterly track due to differences in environmental conditions (PAGASA Climate Data Section, Climatology and Agrometeorology Branch 2011, unpublished data). Accordingly, the rainfall contribution from TCs is also presented in the same quarterly partition.
Figure 5 shows the quarterly total rain, TC rain, and the TC percentage contribution. In the months of DJF (Figs. 5a,e,i), northeasterly winds or the northeast monsoon prevails and brings rain to the eastern coastal regions (Akasaka et al. 2007). Also in the same season, TC rain contributes up to 20% around southwestern Luzon; however, the value of total rainfall in this quarter is only 3% of the annual rain or an average of 23 mm month−1. MAM (Figs. 5b,f,j) marks the transition between the northeast and southwest monsoon regimes. During this boreal spring transition, most convection stays near and south of the equator (Chang et al. 2005) with prevailing easterlies from the Pacific Ocean. Similar to DJF, the TC contribution in MAM shows the same distribution but small total seasonal rainfall (10% of annual rain). The southwest monsoon season usually starts in the second half of May and ends in September (Moron et al. 2009); monsoon winds bring in warm moist air from the South China Sea to the western Philippines (Lagmay et al. 2015). A study by Asuncion and Jose (1980) showed that 43% of the average total annual precipitation in the Philippines falls during this season. The southwest monsoon season also coincides with peak TC occurrence in the months of July and August.
Daily mean (a) total rain, (b) TC rain, and (c) TC percentage contribution (TC/total) from 1951 to 2014. The color scales runs (a) 0 (dark blue) to 10 (dark red) in increments of 2 mm day−1, (b) 0 (dark blue) to 5 (dark red) in increments of 1 mm day−1, and (c) 0% (dark blue) to 70% (dark red) in increments of 10%.
Citation: Journal of Climate 30, 10; 10.1175/JCLI-D-16-0150.1
A high precipitation amount during JJA is seen in the western section of Luzon (Fig. 5c). The distribution of TC-induced rain (Fig. 5g) is consistent with the findings of Cayanan et al. (2011). A detailed discussion is found in the next section. Last, the southeast propagation of the intertropical convergence zone (ITCZ) in SON results in the shifting of monsoonal winds back to the northeast monsoon by the end of October (Moron et al. 2009). Figure 5l shows that the TC rain percentage contribution along northwest Luzon reaches 66.7% during this quarter, when TC tracks are more southwestward. In addition, a higher TC rain contribution is seen in northern Mindanao.
b. TC enhancement of southwest monsoon
Computed TC-induced rainfall is not only from the immediate rainbands of TCs but also from its influence on the southwesterly winds during the southwest monsoon period in JJA. TCs located to the northeast of the Philippines enhance the moisture flux to the west that produces intense convection and orographic lifting along the windward side of the Cordillera mountain range in northwestern Luzon. Figure 6a shows the mean total column water vapor flux for JJA, and Fig. 6b shows where a TC was inside an area northeast of Luzon bounded by 19°–21°N, 125°–127°E that occurred in the same quarter. The composite image of 44 TCs from 1958 to 2014 with intensities stronger than a tropical storm in Fig. 6b shows a large increase and a change in the direction of moisture flow. A mean increase of 411% zonal water vapor flux to the west of the Philippines from 90.2 to 371 kg m−1 s−1 is derived from JRA-55. Figure 6c shows the moisture flux anomaly of the composite 44-TC case. This interaction with the southwest winds increases rainfall in the northwest region (Cayanan et al. 2011; Lagmay et al. 2015).
Seasonal distribution of (a)–(d) total, (e)–(h) TC rain, and (i)–(l) TC percentage contribution. The color scales are as in Fig. 5.
Citation: Journal of Climate 30, 10; 10.1175/JCLI-D-16-0150.1
The calculated TC rain percentage contribution during JJA in Fig. 7 using a TC influence radius of 2.5° (~277 km) to simulate precipitation only from the rainbands shows a different spatial distribution. Here, the northeast region of Luzon is most affected (up to 23%), whereas a comparably lower TC rain contribution is found along western Luzon. This further supports the selection of a 10° TC influence radius used in this study.
TC Percentage contribution (TC/total rain) in JJA at the 2.5° TC influence radius. The color scale runs from 0% (dark blue) to 25% (dark red) in increments of 5%.
Citation: Journal of Climate 30, 10; 10.1175/JCLI-D-16-0150.1
c. Interannual variability of TC rainfall
In this work, k-means clustering was utilized to group regions of distinct climate subtypes according to monthly rainfall variation based on the current 64-yr precipitation dataset. This type of unsupervised clustering technique looks for the least variation within a cluster and the maximum difference between clusters (Everitt et al. 2001). However, it does not determine the optimum number of clusters or the value of k. Here, k = 4 was chosen.
The resulting climate clusters and their corresponding monthly rainfall variability are shown in Fig. 8. Figure 8a shows clusters 1 and 2 covering the western coast and the remainder of western Luzon and Visayas, respectively. These regions have two pronounced seasons: dry from November to April and wet the rest of the year (Fig. 8b). Cluster 3 covers the central (longitudinal) Philippines, where rainfall is evenly distributed throughout the year, and cluster 4 describes the eastern regions, where there is no dry season but more rain during the winter monsoon period. Table 1 summarizes the total rain, TC rain, and the TC rain percentage contribution and their corresponding standard deviation (std dev) for each of the four clusters. The interannual variability of TC-induced rainfall for each climate cluster is also examined.
(a) Climate clusters and (b) corresponding mean monthly rainfall variation from k-means clustering (cluster 1: blue, cluster 2: red, cluster 3: green, and cluster 4: purple).
Citation: Journal of Climate 30, 10; 10.1175/JCLI-D-16-0150.1
Summary of mean and std dev of total rain, TC rain, and TC percentage contribution (×100%) for all clusters (mm day−1).
Figures 9a–d show the annual total rain and TC rain for each climate cluster. The variability of TC rain in the northwestern Philippines correlates well with the variability of total rain. A correlation coefficient of r = 0.69 for cluster 1 and r = 0.72 for cluster 2, both at the 99% significance level, were found. This implies that TC rain has a significant influence on the overall rainfall variability in this region. However, this relationship was not observed for clusters 3 (r = 0.53) and 4 (r = 0.35). Because TC rain makes a low contribution in those clusters, it was not a significant determinant of total rainfall.
Total rain (black and red) and TC rain (green and purple) from APHRODITE (solid) and TRMM (dashed) for (a)–(d) clusters 1–4 (mm day−1) and trends (2000–14) of TC rain (black dashed lines).
Citation: Journal of Climate 30, 10; 10.1175/JCLI-D-16-0150.1
The mean annual rainfall of clusters 1 and 2 (Fig. 9a,c, respectively) follows a small but significant decreasing trend of −0.2 mm decade−1 (p < 0.05) from 1960 to 2010, consistent with the findings of Cruz et al. (2013). However, all clusters as shown in Fig. 9 show an increasing trend from year 2000 onward, with linear trend lines shown as the dashed line and sharp peaks are seen in the years 2008 and 2011. These peaks are more pronounced for cluster 4 along the eastern Philippines. Anomalous heavy rainfall events that occurred during the months of JFM in those 2 yr were likely the result of very strong trade winds in those years as indicated by a high western Pacific trade wind index (not shown). Maximum precipitation is observed along the eastern region, as it is on the windward side of the prevailing monsoon (Chang et al. 2005; Akasaka et al. 2007; Moron et al. 2009). An increasing trend in the west Pacific trade winds was reported by England et al. (2014) to account for the warming hiatus since 2000. Also, this possibly explains the increasing annual rainfall trend and an unprecedented rainfall amount to the east of the Philippines in the past one and a half decades.
Seasonal and interannual variabilities of rainfall in the Philippines are influenced by the extreme phases of ENSO (Hilario et al. 2009; Lyon et al. 2006; Ropelewski and Halpert 1996). This is seen in the positive peaks in total annual rainfall (Fig. 9) in 1971, 1999, 2008, and 2011—all La Niña years—and negative peaks in 1983, 1987, 1997, and 2010, which were El Niño years. A comparison of the ENSO-3.4 index and total rain yielded a high correlation for clusters where the TC contribution is small. For cluster 1, the ENSO-3.4 index and total rain has a low correlation of r = −0.3, cluster 2 has a moderate correlation of r = −0.58, while cluster 3 has r = −0.74 and cluster 4 has r = −0.69—all at the 95% significance level. Since clusters 1 and 2 are affected by the TC-enhanced southwest monsoon, there is no distinctive reduction in TC rain in El Niño years due to the tendency of TCs to track toward the northwest quadrant of the NWP. This increases TC occurrence to the northeast of Luzon, leading to enhanced monsoon winds. In addition, the low correlation can also be explained by the late seasonal reversal toward the final quarter of an ENSO year (Lyon et al. 2006); therefore, ENSO has a relatively small influence on the annual total and TC rainfall. In contrast, clusters 3 and 4 have a small TC rain contribution. Most of the rainfall in these clusters is produced from non-TC mechanisms (i.e., the interaction between land and trade wind during the northeast monsoon) that are well correlated with ENSO events (Webster and Yang 1992).
TC-induced rainfall variability is highest for clusters located in the northwestern half of the country (clusters 1 and 2). A single or a few TC-induced heavy precipitation events can affect the total annual rainfall of certain years. In 1986, Typhoon Wayne (local name: Miding) made a substantial contribution to TC rain over Luzon and is considered one of the longest-lived tropical cyclones in the NWP basin (Fatjo and Wells 1986). Recently, Typhoon Parma (Pepeng) and Tropical Storm (TS) Ketsana (Ondoy) increased the total rainfall in 2009 (Yumul et al. 2013). In 2012 and 2013, two weeklong heavy precipitation events due to the enhancement of southwesterly winds by several persistent TCs affected the west of central Luzon (Lagmay et al. 2015). The TCs described have resulted in a total accumulated rainfall volume in the upper 95th percentile. In 2002 when TC rain was lowest, most TCs that formed in the NWP basin had northeastward recurvature due to a weak NWPSH steering. This resulted in only two TCs—TS Tapah (Agaton) and Severe Tropical Storm (STS) Noguri (Espada)—to come within 100 km of Philippine landmass. In the southern Philippines, 2011–13 show the highest TC-induced rainfall due to several southerly extreme TC events: Typhoon Washi (Sendong) in 2011; Typhoon Bopha (Pablo) in 2012; and STS Sonamu (Auring), TS Shanshan (Crising), and TS Podul (Zoraida) in 2013.
Increasing trends in TC rainfall are seen in all clusters from 2000, with a higher trend along the eastern region. In clusters 1 and 2, increasing TC rain resulted in the reversal of the previously reported (Cruz et al. 2013; Villafuerte et al. 2014) negative southwest monsoon rainfall trend. The TC rain and TC percentage contribution trends from 2000 to 2014 are summarized in Table 2. Cluster 4 shows the highest increasing trend of 1.95 mm day−1 decade−1 in TC rain, while cluster 2 shows the highest increase in TC percentage contribution. Significant increases (p < 0.05) are observed in all clusters.
Trends in TC rain and TC percentage contribution (×100%) for all clusters since 2000 (decade−1).
This study hypothesizes that changes in TC steering mechanisms—namely, a southerly subtropical steering ridge and stronger environmental flow during SON and December (SOND) in the last two decades—affected TC movement. This leads to TCs shifting to more westward, southerly tracks. Figure 10 shows the decadal southwestward migration of the NWPSH represented by the 5870-m geopotential height at the 500-mb isosurface during SOND for the periods 1975–84 (gray solid line), 1985–94 (gray dashed–dotted line), 1995–2004 (dotted line), 2005–14 (dark dashed line), and 2011–14 (dark solid line). The NWPSH is an important steering mechanism of TCs in the NWP basin (Xiang et al. 2013). TCs move along the periphery of the NWPSH, and its westward extent determines whether a TC recurves northward or moves toward a westerly direction. Also, stronger easterly steering flow (850 mb) is seen from a wind anomaly (vectors) for 2011–14 as compared to the 1980–2010 climatological mean. This may also explain the higher TC landfall occurrences along the southern regions of the Philippines found in the analysis by David et al. (2013). In addition, thermodynamic factors such as an increase in tropospheric humidity (Dai 2006; Chan 2008; McCarthy et al. 2009) and warmer sea surface temperature to the east of the Philippines (Comiso et al. 2015) likely contributed to the precipitation amounts of TCs affecting the Philippines.
NWP SOND mean wind anomaly (850 mb) for 2010–14 relative to 1980–2010 (vectors) and 5870-m geopotential height (500 mb) for 1975–84 (gray solid line), 1985–94 (gray dashed–dotted line), 1995–2004 (dotted line), 2005–14 (dark dashed line), and 2010–14 (dark solid line).
Citation: Journal of Climate 30, 10; 10.1175/JCLI-D-16-0150.1
Most studies about the effects of global warming on TCs focus on TC frequency and intensity. There are a few studies, however, that deal with changes in TC rainfall area. Since the TC rainfall area or the TC rain field is easily observable in satellite images, the TC rain field is often used as a measure of TC size. It can be influenced by a TC’s humidity, vorticity, vertical wind shear, and latitude. The TC rain field is also closely related to the TC wind field (Matyas 2010).
Using TRMM satellite observation and a global circulation model, Lin et al. (2015) showed that the TC rain field is controlled primarily by relative sea surface temperature (SST), or SST over a region compared to the tropical mean SST, whereas the TC rainfall rate increases with absolute SST. Both of these phenomena are observed in the tropical western Pacific Ocean to have significant increasing trends (Dai 2006; Durre et al. 2009; McCarthy et al. 2009). In relation to this, various studies have shown that TC strength is well correlated with SST (Emanuel 2005; Webster et al. 2005; Mei et al. 2015). Comiso et al. (2015) observed that the regional SST to the east of the Philippines, referred to as the warm pool (10°S–10°N, 150°–170°E), has been warming by 0.2° decade−1. Their results also showed a high correlation (r = 0.7) between TC maximum wind speed in the NWP and the warm pool SST. Moreover, the frequency of TCs with maximum wind speed greater than 150 km h−1 has also been increasing with 29.6% of 169 TCs in 1984–93, 30.5% of 141 TCs in 1994–2003, and 41.9% of 160 TCs in 2004–13.
Are we seeing wetter TC conditions in the Philippines due to the warming of global surface temperature? The aforementioned studies on the changes of some dynamic and thermodynamic factors seem to suggest so. The long-term southwest progression (Villafuerte et al. 2014) and a more recent westward extension of the NWPSH (Sui et al. 2007; Zhou et al. 2009) are partly explained by warming SST trends. Warming SST in the warm pool region produces more frequent intense TCs in the tropical western Pacific, and consequently more TCs with intense rain field. Surface warming also leads to higher moisture content in the atmosphere (Durre et al. 2009; McCarthy et al. 2009). Last, a larger TC rainfall area in the NWP basin compared to Atlantic TCs is a manifestation of high relative SST in the NWP region (Bretherton et al. 2004; Lin et al. 2015). But while studies had shown an increase in the destructiveness of tropical cyclones (Emanuel 2005; Webster et al. 2005; Wu et al. 2008) since the 1970s, interdecadal variations of dynamic factors conducive to TC formation and intensification that are coupled to oceanographic cycles should also be considered, as Chan (2008) emphasized. Furthermore, the 16–32-yr variation in TC intensity that Chan (2008) reported started its upward cycle around the year 2000, which coincides with the increase in TC rain contribution. Further investigation is warranted to sufficiently answer whether TC rain is indeed increasing due to climate change. It is a question vital to the mitigation and adaptation efforts geared toward anthropogenic climate change in the Philippines.
4. Summary and conclusions
Tropical cyclones are of great socioeconomic importance to the Philippines. In this study, the contribution of TC-induced rainfall was characterized using combined ground and satellite data to produce a blended 64-yr precipitation dataset. Out of the 1673 TCs analyzed, 1273 TCs or 76% had rainfall contribution on Philippine landmass. The highest TC rain (as much as 1400 mm yr−1) and percentage contribution (54.2%) averaged over the entire study period is found along the northwest coast of Luzon in the northern Philippines. TC rain contribution in this region was mainly due to the influence of TCs located to the northeast of the Philippines on the prevailing southwesterly winds during the summer monsoon season and not directly from the immediate TC rainbands. JJA and SON show the highest TC-induced precipitation, as these coincide with the highest TC occurrence months and a climatological southerly–westward TC track, respectively, that influence precipitation distribution. The southern islands of Mindanao showed the least amount of rain from TCs. Nevertheless, TC rain contribution still reached ~20% in that region during SON.
To examine the interannual variability of TC rain contribution, the Philippines was grouped into separate climate subtypes according to monthly rainfall variability using the unsupervised k-means clustering method. A k-means for k = 4 was used in the cluster analysis. Clusters 1 and 2, which cover western Luzon and northwest Visayas, have the highest TC-induced rain and percentage contribution. On the other hand, cluster 4, which covers the eastern coast, has the lowest. Positive (negative) peaks in total annual rainfall coincide with La Niña (El Niño) years. However, regions with less TC rain contribution have a higher correlation to the ENSO-3.4 index. This leads to the conclusion that ENSO has more influence on non-TC rain.
Increasing trends in both TC rain and TC percentage contribution were seen for the whole Philippines after the 1997–99 ENSO event. These trends are likely due to more westward TC tracks brought about by changes in dynamic and thermodynamic factors. These factors affected TC movement and intensity in the recent one and a half decades. Aside from precipitation, the destructive potential of TCs’ intense winds was most likely also affected by the said changes. A separate paper to describe their impacts is underway (G. Bagtasa 2017, unpublished manuscript). The Philippines is an agricultural country with many rain-fed farmlands and together with a rapidly increasing population, it is important to understand the spatiotemporal characteristics of TC-induced precipitation to assess its implication on water resources management and food security of the country.
Acknowledgments
The author would like to acknowledge the generosity of the Oscar M. Lopez Center for Climate Change Adaptation and Disaster Risk Management Foundation, Inc. (OML Center) for supporting the WHATSUP project (NP2013-07UP1). The author would also like to thank the three reviewers for their valuable comments, which helped to improve the quality of the paper.
REFERENCES
Akasaka, I., W. Morishima, and T. Mikami, 2007: Seasonal march and its spatial difference of rainfall in the Philippines. Int. J. Climatol., 27, 715–7252, doi:10.1002/joc.1428.
Asuncion, J., and A. Jose, 1980: A study of the characteristics of the northeast and southwest monsoons in the Philippines. National Research Council of the Philippines Assisted Project, 49 pp. [Available from Philippine Atmospheric Geophysical and Astronomical Services Administration, Agham Road, Dillman, Quezon City, Metro Manila 1100, Philippines.]
Atkinson, G., 1977: Proposed system for near real time monitoring of global tropical circulation and weather patterns. Preprints, 11th Tech. Conf. on Hurricanes and Tropical Meteorology, Miami Beach, FL, Amer. Meteor. Soc., 645–652.
Breña-Naranjo, J. A., A. Pedrozo-Acuña, O. Pozos-Estrada, S. A. Jimenez-Lopez, and M. R. Lopez-Lopez, 2015: The contribution of tropical cyclones to rainfall in Mexico. Phys. Chem. Earth, 83–84, 111–122, doi:10.1016/j.pce.2015.05.011.
Bretherton, C., M. Peters, and L. Back, 2004: Relationships between water vapor path and precipitation over the tropical oceans. J. Climate, 17, 1517–1528, doi:10.1175/1520-0442(2004)017<1517:RBWVPA>2.0.CO;2.
Camargo, S., and A. Sobel, 2005: Western North Pacific tropical cyclone intensity and ENSO. J. Climate, 18, 2996–3006, doi:10.1175/JCLI3457.1.
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, 89A, 123–139, doi:10.2151/jmsj.2011-A08.
Chadwick, R., I. Boutle, and G. Martin, 2013: Spatial patterns of precipitation change in CMIP5: Why the rich do not get richer in the tropics. J. Climate, 26, 3803–3822, doi:10.1175/JCLI-D-12-00543.1.
Chan, J. C. L., 2000: Tropical cyclone activity over the western North Pacific associated with El Niño and La Niña events. J. Climate, 13, 2960–2972, doi:10.1175/1520-0442(2000)013<2960:TCAOTW>2.0.CO;2.
Chan, J. C. L., 2008: Decadal variations of intense typhoon occurrence in the western North Pacific. Proc. Roy. Soc. London, 464A, 249–272, doi:10.1098/rspa.2007.0183.
Chang, C.-P., Z. Wang, J. McBride, and C.-H. Liu, 2005: Annual cycle of Southeast Asia maritime continent rainfall and the asymmetric monsoon transition. J. Climate, 18, 287–301, doi:10.1175/JCLI-3257.1.
Cinco, T. A., R. G. de Guzman, F. D. Hilario, and D. M. Wilson, 2014: Long-term trends and extremes in observed daily precipitation and near surface air temperature in the Philippines for the period 1951–2010. Atmos. Res., 145–146, 12–26, doi:10.1016/j.atmosres.2014.03.025.
Cinco, T. A., and Coauthors, 2016: Observed trends and impacts of tropical cyclones in the Philippines. Int. J. Climatol., 36, 4638–4650, doi:10.1002/joc.4659.
Comiso, J. C., G. J. Perez, and L. V. Stock, 2015: Enhanced Pacific Ocean sea surface temperature and its relation to Typhoon Haiyan. J. Environ. Sci. Manage., 18 (1), 1–10.
Coronas, J., 1912: The extraordinary drought in the Philippines: October, 1911, to May, 1912. Weather Bureau, Department of the Interior, Government of the Philippine Islands, 19 pp.
Corporal-Lodangco, I., and L. Leslie, 2016: Impacts of ENSO on Philippine tropical cyclone activity. J. Climate, 29, 1877–1897, doi:10.1175/JCLI-D-14-00723.1.
Cruz, F. T., G. T. Narisma, M. Q. Villafuerte, K. U. Chen Chua, and L. M. Olaguera, 2013: A climatological analysis of the southwest monsoon rainfall in the Philippines. Atmos. Res., 122, 609–616, doi:10.1016/j.atmosres.2012.06.010.
Dai, A., 2006: Recent climatology, variability, and trends in global surface humidity. J. Climate, 19, 3589–3606, doi:10.1175/JCLI3816.1.
Dare, R., N. Davidson, and J. McBride, 2012: Tropical cyclone contribution to rainfall over Australia. Mon. Wea. Rev., 140, 3606–3619, doi:10.1175/MWR-D-11-00340.1.
David, C., B. Racoma, J. Gonzales, and M. Clutario, 2013: A manifestation of climate change? A look at Typhoon Yolanda in relation to the historical tropical cyclone archive. Sci. Diliman, 25, 79–86.
Durre, I., C. N. Williams Jr., X. Yin, and R. S. Vose, 2009: Radiosonde-based trends in precipitable water over the Northern Hemisphere: An update. J. Geol. Res., 114, 1–8, doi:10.1029/2008JD010989.
Emanuel, K., 2005: Increasing destructiveness of tropical cyclones over the past 30 years. Nature, 436, 686–688, doi:10.1038/nature03906.
England, M. H., and Coauthors, 2014: Recent intensification of wind-driven circulation in the Pacific and the ongoing warming hiatus. Nat. Climate Change, 4, 222–227, doi:10.1038/nclimate2106.
Everitt, B., S. Landau, and M. Leese, 2001: Cluster Analysis. 4th ed. Oxford University Press, 237 pp.
Fatjo, S. J., and F. H. Wells, Eds., 1986: Annual tropical cyclone report. U.S. Naval Oceanography Command Center, JTWC, 193 pp.
Flores, J., and V. Balagot, 1969: Climate of the Philippines. Climates of Northern and Eastern Asia, H. Arakawa, Ed., World Survey of Climatology, Vol. 8. Elsevier, 159–213 pp.
Hilario, F., R. de Guzman, D. Ortega, and P. Hayman, 2009: El Niño southern oscillation in the Philippines: Impacts, forecasts, and risk management. Philipp. J. Dev., 36, 9–34.
Jiang, H., J. B. Halverson, J. Simpson, and E. Zipser, 2008: Hurricane “rain potential” derived from satellite observations aids overland rain prediction. J. Appl. Meteor. Climatol., 47, 944–959, doi:10.1175/2007JAMC1619.1.
JMA, 2013: JRA-55: Japanese 55-year Reanalysis, daily 3-hourly and 6-hourly data. National Center for Atmospheric Research Computational and Information Systems Laboratory Research Data Archive, accessed 6 June 2016, doi:10.5065/D6HH6H41.
Kubota, H., and B. Wang, 2009: How much do tropical cyclones affect seasonal and interannual rainfall variability over the western North Pacific? J. Climate, 22, 5495–5510, doi:10.1175/2009JCLI2646.1.
Lagmay, A., G. Bagtasa, I. Crisologo, B. Racoma, and C. David, 2015: Volcanoes magnify Metro Manila’s southwest monsoon rains and lethal floods. Front. Earth Sci., 2, 1–9, doi:10.3389/feart.2014.00036.
Lin, Y., M. Zhao, and M. Zhang, 2015: Tropical cyclone rainfall area controlled by relative sea surface temperature. Nat. Commun., 6, 6591, doi:10.1038/ncomms7591.
Liu, J., B. Wang, M. Cane, S. Yim, and J. Lee, 2013: Divergent global precipitation changes induced by natural versus anthropogenic forcing. Nature, 493, 656–659, doi:10.1038/nature11784.
Liu, Z., 2015: Evaluation of precipitation climatology derived from TRMM Multi-Satellite Precipitation Analysis (TMPA) monthly product over land with two gauge-based products. Climate, 3, 964–982, doi:10.3390/cli3040964.
Lyon, B., and S. Camargo, 2009: The seasonally-varying influence of ENSO on rainfall and tropical cyclone activity in the Philippines. Climate Dyn., 32, 125–141, doi:10.1007/s00382-008-0380-z.
Lyon, B., H. Cristi, E. Verceles, F. Hilario, and R. Abastillas, 2006: Seasonal reversal of the ENSO rainfall signal in the Philippines. J. Geophys. Res., 33, L24710, doi:10.1029/2006GL028182.
Matyas, C., 2010: Associations between the size of hurricane rain fields at landfall and their surrounding environments. Meteor. Atmos. Phys., 106, 135–148, doi:10.1007/s00703-009-0056-1.
McCarthy, M., P. Thorne, and H. Titchner, 2009: An analysis of tropospheric humidity trends from radiosondes. J. Climate, 22, 5820–5838, doi:10.1175/2009JCLI2879.1.
Mei, W., S. Xie, F. Primeau, J. McWilliams, and C. Pasquero, 2015: Northwestern Pacific typhoon intensity controlled by changes in ocean temperatures. Sci. Adv., 1, e1500014, doi:10.1126/sciadv.1500014.
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, doi:10.1007/s00382-008-0520-5.
Ren, F., G. Wu, W. Dong, X. Wang, Y. Wang, W. Ai, and W. Li, 2006: Changes in tropical cyclone precipitation over China. Geophys. Res. Lett., 33, L20702, doi:10.1029/2006GL027951.
Ribera, P., R. Garcia-Herrera, and L. Gimeno, 2008: Historical deadly typhoons in the Philippines. Weather, 63, 194–199, doi:10.1002/wea.275.
Ropelewski, C., and M. Halpert, 1996: Quantifying Southern Oscillation–precipitation relationships. J. Climate, 9, 1043–1059, doi:10.1175/1520-0442(1996)009<1043:QSOPR>2.0.CO;2.
Sui, C., P. Chung, and T. Li, 2007: Interannual and interdecadal variability of the summertime western north Pacific subtropical high. Geophys. Res. Lett., 34, L11701, doi:10.1029/2006GL029204.
Villafuerte, M., J. Matsumoto, I. Akasaka, H. Takahashi, H. Kubota, and T. Cinco, 2014: Long-term trends and variability of rainfall extremes in the Philippines. Atmos. Res., 137, 1–13, doi:10.1016/j.atmosres.2013.09.021.
Wang, B., and J. C. L. Chan, 2002: How strong ENSO events affect tropical storm activity over the western North Pacific. J. Climate, 15, 1643–1365, doi:10.1175/1520-0442(2002)015<1643:HSEEAT>2.0.CO;2.
Wang, B., J. Liu, H. J. Kim, P. J. Webster, and S. Y. Yim, 2012: Recent change of the global monsoon precipitation (1979–2008). Climate Dyn., 39, 1123–1135, doi:10.1007/s00382-011-1266-z.
Wang, J. J., R. F. Adler, G. J. Huffman, and D. Bolvin, 2014: An updated TRMM composite climatology of tropical rainfall and its validation. J. Climate, 27, 273–284, doi:10.1175/JCLI-D-13-00331.1.
Webster, P. J., and S. Yang, 1992: Monsoon and ENSO: Selective interactive systems. Quart. J. Roy. Meteor. Soc., 118, 877–926, doi:10.1002/qj.49711850705.
Webster, P. J., G. Holland, J. Curry, and H. Chang, 2005: Changes in tropical cyclone number, duration and intensity in a warming environment. Science, 309, 1844–1846, doi:10.1126/science.1116448.
Weinkle, J., R. Maue, and R. Pielke, 2012: Historical global tropical cyclone landfalls. J. Climate, 25, 4729–4735, doi:10.1175/JCLI-D-11-00719.1.
World Bank, 2015: Employment in agriculture (% of total employment). International Labour Organization. Subset used: Philippines, accessed 16 February 2016. [Available online at http://data.worldbank.org/indicator/SL.AGR.EMPL.ZS.]
Wu, L., B. Wang, and S. A. Brauna, 2008: Implications of tropical cyclone power dissipation index. Int. J. Climatol., 28, 727–731, doi:10.1002/joc.1573.
Xiang, B., B. Wang, W. Yu, and S. Xu, 2013: How can anomalous western North Pacific Subtropical High intensify in late summer? Geophys. Res. Lett., 40, 2349–2354, doi:10.1002/grl.50431.
Yasutomi, N., A. Hamada, and A. Yatagai, 2011: Development of a long-term daily gridded temperature dataset and its application to rain/snow discrimination of daily precipitation. Global Environ. Res., 15, 165–172.
Yatagai, A., A. Kitoh, K. Kamiguchi, O. Arakawa, and A. Hamada, 2012: APHRODITE: Constructing a long-term daily gridded precipitation dataset for Asia based on a dense network of rain gauges. Bull. Amer. Meteor. Soc., 93, 1401–1415, doi:10.1175/BAMS-D-11-00122.1.
Yumul, G., C. Dimalanta, N. Servando, and N. Cruz, 2013: Abnormal weather events in 2009, increased precipitation and disastrous impacts in the Philippines. Climatic Change, 118, 715–727, doi:10.1007/s10584-012-0661-8.
Zhang, L., and T. Zhou, 2011: An assessment of monsoon precipitation changes during 1901–2001. Climate Dyn., 37, 279–296, doi:10.1007/s00382-011-0993-5.
Zhou, T., and Coauthors, 2009: Why the western Pacific subtropical high has extended westward since the late 1970s. J. Climate, 22, 2199–2215, doi:10.1175/2008JCLI2527.1.
