T.A. Cinco et al. / Atmospheric Research 145–146 (2014) 12–26
Table 2
Extreme temperature trend values for hot days and cold nights from
synoptic weather stations in the Philippines.
Stations
Freq.
Hot days
p (95%)
Trend
Freq.
Alabat
Ambulong
Baguio city
Cabanatuan
Calapan
Casiguran
Catarman
Catbalogan
Coron
Cuyo
Daet
Dagupan
Davao
Dipolog
General Santos
Hinatuan
Iloilo
Infanta
Laoag
Legaspi
Malaybalay
Masbate
Puerto Princesa
Romblon
Roxas
Science Garden Q.C)
Surigao
Tacloban
Virac (synoptic)
Zamboanga
-0.2004
-0.0097
0.2202
0.2044
0.1116
0.0702
0.2837
0.3923
0.2142
0.5355
0.0067
0.0406
-0.0218
-0.1106
0.1119
0.1263
0.2537
0.3102
0.0959
-0.0691
0.4663
0.1163
0.1068
0.0014
-0.2455
0.0449
0.2064
0.5685
-0.2471
0.6163
0.0016
0.1379
0.0000
0.0040
0.0012
0.6896
0.0000
0.0000
0.0002
0.0000
0.8859
0.2520
0.4359
0.0014
0.0628
0.0102
0.0000
0.0000
0.0059
0.0200
0.0000
0.3397
0.0889
0.6589
0.0042
0.6876
0.0000
0.0000
0.0027
0.0000
<
<
>
>
>
>
>
>
>
>
>
>
<
<
>
>
>
>
>
<
>
>
>
>
<
>
>
>
<
>
0.0655
-0.0186
0.0140
-0.0499
0.0171
-0.0862
-0.1465
-0.0286
-0.2330
-0.1876
-0.1173
0.0820
-0.1428
0.6713
-0.3286
-0.0956
0.0155
-0.0569
-0.0481
-0.0038
0.0302
-0.0183
-0.0013
-0.1047
-0.0284
-0.1488
-0.1064
0.0114
0.0028
-0.1548
Cold nights
p (95%)
Trend
0.3450
0.1702
0.3695
0.0163
0.8276
0.0025
0.0000
0.0613
0.0639
0.0000
0.0000
0.0121
0.0000
0.4828
0.0000
0.0000
0.6674
0.0112
0.0093
0.0318
0.8044
0.2264
0.7771
0.0000
0.0061
0.0000
0.0000
0.4545
0.6815
0.0000
>
<
>
<
>
<
<
<
<
<
<
>
<
>
<
<
>
<
<
<
>
<
<
<
<
<
<
>
>
<
station at Baguio located in a mountainous area of the north
western Philippines, exhibiting a significant increase in
rainfall intensity similar to other lowland stations in the
region (Laoag and Infanta). Overall, only six of the stations
produced negative trends in rainfall intensity, all of which
were statistically insignificant with the remainder showing
positive but similarly insignificant trends.
The majority of the stations showed positive trends in the
frequency of severe rainfall events with only four stations
showing negative trends all of which are statistically insignificant. The seven stations showing significant increases are
again distributed throughout the islands of the Philippines
from the far north to the far south and from east to west.
Stations at Calapan and Casiguran showed significant increases
in frequency whilst those at Iloilo, Tacloban (both within the
central island group of the Philippines and recently affected by
typhoon Haiyan), Laoag, Infanta and Cotabato all showed a
significant increase in both intensity and frequency.
The persistent level of variability between stations and
the lack of a discernible, coherent trend in the data may,
in part be due to the analysis not being sensitive to the
inter seasonal variation, different orographic conditions and
because there are complex, localised oceanic–atmospheric
interactions at play in an archipelagic nation such as the
Philippines. Indeed, this is consistent with the reference studies
which have found little evidence of spatial coherence when
considering extreme rainfall (Griffiths et al., 2005; Manton et
al., 2001). A weaker signal for rainfall variables could also be
associated with the higher inter-annual variability of extreme
rainfall events as well as the comparatively shorter timescale
21
over which change occurs when compared to that of temperature variables.
4.3. Results in a national and regional context
Rainfall trends for both intensity and frequency of extreme
events for the whole of the Philippines based on observed data
are in the majority not statistically significant. However, some
stations do show a significant increase in both intensity and
frequency and some of these frequently experience the effects
of tropical cyclones including typhoons. More coherent at the
national level are the extreme daily temperature trends for cold
nights and hot days. The statistically significant trends in
extreme events when considered in conjunction with the
observed warming trends described earlier combine to reveal a
warming climate over the last 30 years.
These findings at a national level appear to be broadly
consistent with both previous regional studies for similar
temperature indicators for the Asia-Pacific region and with
global studies. One of the earliest and most comprehensive
studies to which the findings of this study can be compared is
that of Manton et al. (2001) which used data from five
stations in the Philippines (Basco, Baguio, Daet, Dumaguete
and Tuguegarao) considered to have the highest quality time
series data for the period 1961–1998. The data common to
both that study and this is from the stations at Baguio and
Dumaguete. Analysis of the data from the other stations used
by Manton et al. (2001) available for the period 1951–2010
was deemed unreliable. This makes a direct comparison of
results from those studies and this one difficult, not least
because station level information is not presented in detail in
earlier studies (although Manton et al. (2001) noted that
there was a significant increase in hot days at the Baguio
station which is consistent with this study). Indeed, the aim
of this study was not to conduct a direct comparison of
national level data with that of Manton et al. (2001). Instead
this study uses the same methodology employed by Manton
et al. (2001) to provide an updated, contemporary insight
into the temperature and precipitation trends within the
Philippines, providing a more detailed analysis from a greater
number of stations and presenting the results in a regional
and global context.
At the regional level in Southeast Asia, the findings here are
broadly consistent with other earlier studies demonstrating a
largely coherent warming trend over much of the region with a
few exceptions found in Australia and the southern Pacific
(Griffiths et al., 2005). Sub-nationally, the station at Baguio was
found to have a large, statistically significant increase in hot
days, which is in agreement with this study. Griffiths et al.
(2005) also established a link between mean temperature
trends and extreme events, particularly in the tropical Pacific
region, including the Philippines, suggesting that even small
increases in mean temperatures may lead to an increased
probability of extreme events. Given that this study has
highlighted the increasing trends of mean, minimum and
maximum temperatures observed at a national level, the link
between extreme events and these indicators for a nation
vulnerable to the impacts of climate change is notable.
Manton et al. (2001) observed an increased frequency in
hot extremes (hot days) and decrease in cold extremes (cold
days) across the Asia Pacific region as well as the notable