The variability of the sea surface temperature (SST) in
the northwest Pacific has been studied on seasonal, annual and interannual
scales based on the monthly datasets of extended reconstructed
sea surface temperature (ERSST) 3b (1854–2017, 164 years) and
optimum interpolation sea surface temperature version 2 (OISST V2 (1988–2017,
30 years). The overall trends, spatial–temporal
distribution characteristics, regional differences in seasonal trends and
seasonal differences of SST in the northwest Pacific have been calculated
over the past 164 years based on these datasets. In the past 164 years, the
SST in the northwest Pacific has been increasing linearly year by year, with
a trend of 0.033
The ocean is one of the important components of the ocean–atmosphere
coupling system (Chelton and Xie, 2010; Wu et al., 2019a, b, 2020). Relative
to the atmosphere, the ocean has characteristics such as slow change and
large heat capacity (England et al., 2014). Because of the gradual changes
in the ocean, climate change at the interannual, decadal and longer
timescales may be closely related to the ocean (Trenberth and Hurrell, 1994;
Ault et al., 2009). Sea surface temperature (SST) is the basis for the
interaction between the ocean and the atmosphere (Wu et al., 2019c, d), and
it characterizes the combined results of ocean heat content (Buckley et al.,
2014; Griffies et al., 2015) and dynamic processes (Takakura et al., 2018). It
is a very important parameter for climate change and ocean dynamics processes, and
reflects sea–air heat and water vapor exchange. Observations and numerical
simulations show that large-scale sea surface temperature anomalies of over
20
Bathymetric map of the northwest Pacific and ocean circulation.
The northwest Pacific is particularly affected by the El Niño in the eastern Pacific and determines the oceanic climate change in China (Hu et al., 2018). On one hand, climate change causes an increasing SST in the northwestern Pacific, which increases the vertical stratification of the water, affects the atmospheric circulation and changes the intensity and period of coastal winds and upwelling. On the other hand, the 10-year periods of the Pacific Decadal Oscillation (PDO) and the El Niño–Southern Oscillation (ENSO) occur on average every 2 to 7 years, resulting in large variations in upwelling (Xiao et al., 2015; Yang et al., 2017; Xue et al., 2018). These factors will all lead to the impact on the marine environment in Chinese coastal areas, causing land-based droughts, floods and climate disasters (Xu et al., 2018). Therefore, it is very urgent to study the impact of climate change on SST in the northwest Pacific and China offshore. As one of the main parameters of global climate change and one of the important characterizations and predictors of El Niño, the study of SST changes is particularly important.
Previous scholars have done a lot of work on the changing trend of SST.
According to the Fifth Assessment Report (AR5) of the Intergovernmental
Panel on Climate Change (IPCC), the global SST warming trend was 0.064
So far, two types of main meteorological SST datasets have been obtained:
one based on measured mid-resolution (1–5
Satellite remote sensing can achieve large-area simultaneous measurements with high temporal and spatial resolution. The remote sensing SST obtained is conducive to a more comprehensive and rapid understanding of oceanographic phenomena that affect the ocean surface, including El Niño (Robinson, 2016). At present, about 30 years of satellite remote sensing SST data have been accumulated (Franch et al., 2017), and a set of sea surface temperature data has been provided to study the conditions for the occurrence and development of ocean surface heat change modes in the temporal and spatial span and resolution. So, satellite remote sensing SST has received widespread attention in recent years.
Atlantic Multidecadal Oscillation (AMO) index
At present, based on satellite remote sensing data, the timescales for the study of changes in SST in the northwest Pacific, especially in China offshore, are mostly within 20 years, which is relatively short for studying climate change (Song et al., 2018; Pan et al., 2018). Most of the research is targeted at specific local sea areas, and there is less research on the changes of the SST in the northwest Pacific covering all marginal seas of China. Therefore, it is necessary to study the SST variation of large-scale and long-term sequences based on satellite remote sensing data.
Previous scholars have made great contributions to the study of global
warming, but most of them are the overall changes in the regional average
SST, and they tend to ignore the characteristics of changes in certain key
sea areas. There are great differences in the trends of SST in different sea
areas. The long-term trends of the SST changes in the northwest Pacific
(0–60
High-spatial-resolution SST datasets including average SST field and monthly SST anomaly (SSTA) field have been obtained. In view of the fact that there are many interannual and intra-annual changes, this paper analyzes the characteristics of SST changes based on these datasets. The trend, the interdecadal changes in SST and their causes, and the correlations with the climate parameters and indices such as NINO3.4 index are relatively low. The ocean thermal dynamic phenomenon is preliminarily discussed. The datasets are processed and analyzed to study the trend changes of SST in the northwest Pacific, to explore the correlation and response mechanisms with climate systems such as ENSO and PDO, and to conduct a detailed analysis of typical sea areas.
The northwest Pacific is the northwest region of the Pacific, defined as
the offshore region of 0–60
The temporal variability of annual SST.
Variability of seasonal/annual SST.
Several data sources are used to analyze the long-term temporal and spatial
variability of SST in the northwest Pacific in this present study. Long-term
statistics are based on the monthly SST data from ERSST 3b (1854–2017) (Smith et al., 2008). The
ERSST dataset is a global monthly sea surface temperature analysis derived
from the International Comprehensive Ocean–Atmosphere Dataset with missing
data filled in by statistical methods. This monthly analysis begins in
January 1854 and continues to the present (
The seasonal mean data are obtained by averaging the monthly average SST after the above-mentioned processing. The spring is March, April and May (MAM), the summer is June, July and August (JJA), the autumn is September, October and November (SON), and the winter is December of the previous year and January and February (DJF).
The SST anomaly is the deviation from the long-term SST average of the observations of the SST describing a particular area and time. The year anomaly represents the deviation of the average of the SST for a given year from the mean of the multi-year SST. The month anomaly represents the deviation of the average of the SST for a particular month from the average of the SST for that particular month for many years. In this paper, the mean value from 1854 to 2017 is taken as the climate mean state, and the sea surface temperature anomaly is subtracted from the SST field to obtain the SSTA field.
The average trend of SST (unit:
NWP: northwest Pacific; NH: Northern Hemisphere; GLO: globe. All the trends are significant at the 95 % confidence level.
The Atlantic Multidecadal Oscillation (AMO) is a climate cycle that affects
the SST of the North Atlantic Ocean based on
different modes on multidecadal timescales
(
The correlation between the SST and the atmospheric parameters is analyzed
based on the ERA-Interim data. ERA-Interim refers to the European Centre for
Medium-Range Weather Forecasts (ECMWF), which is an independent
intergovernmental organization supported by 34 countries. Its goal is to
develop numerical methods for midterm weather forecasting. The country
provides forecasting services, conducts scientific and technological
research to accumulate forecasts, and accumulates meteorological data.
ERA-Interim is the latest global reanalysis product developed by ECMWF. The
weather data and climate data from January 1988 to December 2017 are used in
this paper, such as sea surface temperature, sea-to-air interface heat flux
and wind field data at a height of 10 m; the spatial resolution of these
datasets is
Regression analysis is an important part of mathematical statistics and multivariate statistics. It is a mathematical method to study the correlation between variables and variables. The regression analysis has a wide range of applications in the statistical forecasting of oceans and atmospheres. It is used to analyze the statistical relationship between a variable (called forecast) and one or more independent variables (called predictions) and to establish a forecast. The regression equation is produced by the quantity and forecast factor, and then based on this equation to make predictions of the forecast volume. Regression analysis includes linear regression and non-linear regression. The linear regression is commonly used, and a linear regression analysis method is used in this paper.
Use
The NINO3.4 index and SST/SSTA during 1988 to 2017. (El Niño is in pink and La Niña is in blue.)
For the observation data
Spatial distribution of monthly SST over the 1988–2017 period.
Spatial distribution of seasonal/annual SST over the
1988–2017 period:
With the gradual warming of the global climate, the average temperature of the ocean is also rising. In order to reflect the overall trend of SST in the northwest Pacific over the past 164 years (1854–2017), the average monthly SST data from 1854 to 2017 were used. The time series curve of SST in the northwest Pacific, the Northern Hemisphere and the global ocean was obtained by processing, and the overall trend of the SST was analyzed, as shown in Fig. 3. As can be seen from the figure, SSTs in the different regions have shown an increasing trend and SST has shown a significant increasing trend since the 20th century.
The SST datasets were used to calculate the SST anomaly time series and its
linear variation trend in the northwest Pacific, the Northern Hemisphere and
the global ocean, as shown in Fig. 3. The slope of the linear equation with
one unknown obtained by least-squares fitting is the annual change rate of
SST, as shown in Table 1. It shows the increasing trend of SST at different
timescales. It can be seen that the data show that the SST in the
different regions has shown a significant warming trend as a whole. It can be
seen from Table 1 that, from 1854 to 2017, the SST trend of northwest
Pacific, Northern Hemisphere and global ocean has increased by 0.033 to 0.035
There exist decadal to multidecadal variations in the SST and SST anomalies series, with a general cool period from the 1880s to the 1910s, a weak warm period from the 1920s to the 1940s, a weak cool period from the 1970s to the 1980s and a recent warm period from the 1990s to the present. Figure 3 also shows that the interannual to decadal variability is larger in the northwestern Pacific, and it is smaller in the global ocean, indicating an increase in SST anomaly variability with the area. It is also interesting to note that the last 10 years see a larger increasing trend of annual mean SST than that for the last 164, 100, 50 and 30 years, indicating an obvious speed-up of warming of the northwest Pacific, Northern Hemisphere and global ocean occurs in the last 10 years, and the growth rate over the past decade has been around 10 times that of the past 164 years.
In the past 164 years, the correlation coefficient of SST trends in the
northwest Pacific was 0.73. It passed the 95 % significance test, which
shows that the linear trend is significant, and the regression coefficient
is 0.0033. This shows that, in the past 164 years, the SST in the northwest
Pacific has been increasing linearly year by year at a rate of
0.033
In order to demonstrate the seasonal variation of the SST trend in the
northwest Pacific, the SST at
Figure 4a and b show seasonal and annual mean SST and SST anomalies
series. The blue lines are their trends of every seasonal mean SST and SST
anomaly series for the western Pacific during 1854–2017; the red lines are
their trends during 1988–2017. The increasing trends during 1854–2017 are
between 0.032 and 0.035
An El Niño or La Niña event is identified if the NINO3.4 index
exceeds
The correlation coefficient between SST and the atmospheric components (level of significance equal to 0.05).
In the analysis of the SST changes in the northwest Pacific during the past 164 years, it has been found that there was a strong warming trend in SST over the past 30 years since 1988. It had been shown that the SST in the northwest Pacific had an overall warming trend starting from the 1970s in the previous studies (Zhou et al., 2009; Kosaka and Xie, 2013) and this study. The time series of the SST in the northwest Pacific from 1988 to 2017 was plotted as shown in Fig. 4c.
Yamamoto's (1986) method has been used to determine the extremum point, and
the formula is
The monthly average sea surface temperature in the northwest Pacific is
represented by an undulating curve, as shown by the dashed blue line in Fig. 5,
and the sea surface temperature anomaly is a dotted red line. The
positive value is filled in yellow, and the negative value is filled in
cyan. The NINO3.4 index is one of several ENSO indicators based on sea surface temperatures. NINO3.4 is the average
sea surface temperature anomaly in the region bounded by 5
Study regions defined in this paper.
Annual
It can be seen from Fig. 5 that the SSTA minimum value point occurs from 1989 to 1996; the maximum value point occurs in 1998 and 2016, and the maximum year coincides with the El Niño year. It is shown that the anomalous changes of the SST in the northwest Pacific are closely related to the occurrence year of ENSO. The changes of the SST in the northwest Pacific are obviously affected by the anomalous changes of SST in the equatorial Pacific. The average SSTA was basically negative before 1996 and the basic value after it was positive. That is, the average SSTA was generally lower than the average of 1988–2017 before 1996, and the average SSTA after 1996 was basically higher than the average of 1988–2017, which is also reflected in Fig. 4c.
Long-term monthly mean SST of the marginal seas of China
during 2008–2017:
Annual and seasonal SST characteristics of the study area in China offshore based on monthly data from 1988 to 2017.
Peak value and time of the annual and seasonal SST of the study area in China offshore based on monthly data from 1988 to 2017.
Figure 6 shows the spatial distribution of the 30-year average SST for each
month of 1988–2017. From the figure, we can find that the spatial
distribution of annual average SST in each month is similar, and the SST is
higher in the low-latitude (near the Equator) region and lower in the
high-latitude region. In the low-latitude region, SST is more evenly distributed
along the latitudes in January to April and November to December, and is
higher in the south and lower in the north. From May to October, the
distribution of SST along the latitude is tilted, showing the distribution
characteristics as higher in the southwest and lower in the northeast, which
is affected by the ocean circulation. In addition, as can also be seen in
Fig. 6, in the low-latitude region, the SST range of change in different
months is relatively small, between 27 and 33
Figure 7 shows the spatial distribution of seasonal and annual mean SSTs during
the 1988–2017 period. As can be seen from the figure, the spatial
distribution of average SST in each season and annually is similar, and
similar to the monthly results (Fig. 6). In the low-latitude region, the SST
is higher but in the high latitudes. SST is relatively low. Annual mean SST
decreases with increasing latitude, with high temperature ranging from
26 to 28
Figure 8 shows the results of SST anomaly in three characteristic stages. Figure 8a shows the SST anomaly for the annual 1998 minus 1988–2017, Fig. 8b is the annual SST difference between the 10 years after 1998 (1998–2007) and the previous 10 years (1988–1997), and Fig. 8c is the SST anomaly for the last 10 years (2008–2017) and the past 30 years (1988–2017).
It can be seen that there was a significant positive anomaly across the past
30-year average in 1998 from Fig. 8a. The positive anomalies around
1.0
It can be seen from Fig. 8b that the SST during the 10 years from 1998 to
2007 has significantly increased compared with the previous 10 years from
1988 to 1997. The positive anomaly is 0.4 to
0.8
Figure 8c shows the anomalous results of SST over the last 10 years (2008–2017) and relatively nearly 30 years (1988–2017). As can be seen from the figure, in addition to the Bohai Sea, the Yellow Sea and the southern region of Japan, there is a wide range of positive anomaly in other regions, and the past 10 years have increased on average in the past 30 years. From Fig. 4a and b, we have known that the increasing trend of SST over the past 30 years is around 3–4 times that of the rising trend of SST over the past 164 years. Therefore, the increasing trend of SST in the past 10 years is more significant, which is consistent with the results in Fig. 4c and Table 1.
Based on monthly data from ERA-Interim, there is some correlation between SST and atmospheric parameters that has been shown in Fig. 9; all marked patterns are at the level of significance equal to 0.05. It can be seen from Fig. 9a that there is a non-significant correlation between SST and NAO but in China offshore and around the region. It shows a weak negative correlation between China offshore SST and NAO. The PDO is an important factor of climate change of the northwest Pacific, and it has a strong correlation with ENSO. The PDO has a great influence on the Asian monsoon and climate change in the northwest Pacific and is closely related to ENSO. The significant negative correlation between SST and PDO can be seen in Fig. 9b. The NINO3.4 index is usually used to indicate the intensity of the El Niño/La Niña events. So there is a significant negative correlation between SST and the NINO3.4 atmospheric parameter in Fig. 9d.
There is a significant positive correlation between SST and the Southern
Oscillation Index (SOI) in Fig. 9c, which is a standardized index based on
the observed sea level pressure differences between Tahiti and Darwin,
Australia. The monthly correlation between SST and temperature at 2 m (T2) is high throughout the
study region, most markedly (
The maximum negative correlation between the effect of 10 m wind speed (WS10) on SST occurs in the southeastern northwest Pacific and is significant only in a small region. However, the direct correlation between V10 and SST is significant and positive over more of the northwest Pacific.
China offshore is defined as the four sea areas of the Bohai Sea, Yellow
Sea, East China Sea and South China Sea, and includes the Kuroshio
Extension, the part of northwest Pacific and the sea surrounding Japan in
this study, which is defined as the offshore region of 5–41
Figure 11 shows the spatial distribution of seasonal and annual mean SSTs in China offshore during the 1988–2017 period. Annual mean SST decreases with
increasing latitude, with high temperature ranging from 26 to
28
The monthly mean surface temperature changes over the past 10 years in the
three regions (BYS, ECS and SCS) and the whole sea area (China offshore) are
shown in Fig. 12. Figure 12a shows the year-by-year variation of SST in
different regions in the last 10 years, and Fig. 12b shows the monthly SST
variations in different regions in the past 10 years. The change variability
of SST in different regions is basically synchronized. The minimum
temperature basically occurs in February and the warmest occurs in August.
The fluctuation range of SST in BYS is the largest, basically between 5 and 22
Table 2 shows the annual and seasonal SST characteristics of the study area in China offshore based on monthly data from 1988 to 2017. It can be found that
in addition to the winter and spring in the BYS, the SST in each season of
other regions shows an increasing trend from the table. The average increasing
trend of SST during 1988 to 2017 in BYS is 0.015
The northwest Pacific sea surface variability is affected by a combination of oceanic and atmospheric processes and displays significant regional and seasonal behavior. Monthly SST datasets based on ERSST 3b (1854–2017, 164 years) and OISST V2 (1988–2017, 30 years) are used to make some long-term temporal and spatial variability statistics. The following conclusions can be drawn from the analysis.
In the last 164 years, SST in the northwest has gradually increased, with an
increasing trend of 0.033
From the perspective of spatial distribution, the annual mean SST decreases
with increasing latitude, with high temperatures ranging from 27
to 33
There are many correlations between the SST and some climate indices and atmospheric parameters, such as PDO, SOI, NINO3.4, total water vapor column (TCW), T2, sea level pressure (SLP), PRCP and wind speed at 10 m (U10, V10 and WS10). Two very significant positive correlations between SST and T2, TCW have been found, of which the correlation coefficient between SST and T2 exceeded 98 %. PDO and NINO3.4 are negatively correlated with SST, and the correlation between other indices and parameters and SST is weak.
The whole China offshore area was divided into three sections to analyze its
spatial variability in different regions, which are the BYS, ECS and SCS. The SST in the
BYS is coolest, with a range from 5 to 22
Most climate data and figures were sourced from Climate Reanalyzer (
All co-authors were responsible for data collection. ZW and JC wrote the paper, and all co-authors discussed results and assisted with writing.
The authors declare that they have no conflict of interest.
The authors would like to thank anonymous reviewers and the handling topic editor, Dr. Neil Wells.
This research has been supported by the National Natural Science Foundation of China (grant nos. 51839002, 51809023 and 51879015). Partial support also comes from the Research Foundation of Education Bureau of Hunan Province, China (grant no. 19C0092).
This paper was edited by Neil Wells and reviewed by one anonymous referee.