Global warming is the result of increased carbon dioxide and other greenhouse gases concentrations, and it is likely to have some climatic and weather hazards on the eco-environments worldwide [1
], leading to losses of property and human lives [5
] due to floods, droughts, and cyclones. Climate change is largely expressed by temperature and precipitation variability, and extreme precipitation intensity will continue to strengthen as global warming continues [6
]. According to the investigated results of the Intergovernmental Panel on Climate Change [9
], of particular importance are possible changes in extreme events over large parts of the world, including the intensity, frequency, timing, and extent of those climate-related extremes.
Some evaluations of numerical climate model simulations and observational records revealed that the annual average precipitation was increasing with increasing mean temperature at regional and global scales [2
]. Zhang et al. [12
] used the Soil and Water Assessment Tool (SWAT) and utilized global climate model simulations (three scenarios) to forecast the future climate conditions for 2013–2042; they suggested that there is an increasing trend in annual rainfalls under all scenarios and pronounced warming trends in temperature. However, Pierre and Ayan [13
] noted that precipitation decreased with increasing temperature in Djibouti city from 1966 to 2011.
Additionally, the fact that variability of climatic extremes indices is more sensitive than the mean values of climate change [6
] has been universally acknowledged by the scientific community. Studies showed that increasing temperatures and/or precipitations can strengthen the climatic extremes [15
], which could possibly lead to changes in precipitation frequency, amount, and intensity; river flow; soil moisture content; and evapotranspiration rate on regional and global scales [1
]. For example, the increase of precipitation was due to the increase of the rain frequency and intensity in the Chinese Tianshan Mountains from 1961 to 2011 [19
], the heavy precipitation frequency increased with significant increasing temperature [17
], the heavy and sustained precipitation events led to waterlogging, landslides, and some other disasters over part of central-eastern China, resulting in severe damage to crops in the lower reaches of the Yangtze River [20
More and more evidence from climate models shows that most temperature indices that are derived from daily minimum temperature are changed as a result of warming [21
], and the trend of daily minimum temperature warming has a greater impact than that from maximum temperature [24
]. For example, in the Djibouti City of eastern Africa, extreme temperatures (≥45 °C) in the past decade were 15 times more frequent than in the period 1966–1975, while extremely cold nights with a minimum temperature equal to or lower than 8.6 °C disappeared during 1966–2011 [13
]. A study in Poland also indicated that fifteen out of sixteen of the warmest years on record (1880–2015) occurred since 2001. This means that each of the past fifteen years since 2001 has been warmer by at least 0.54 °C than the long-term average from 1910 to 2000 [25
]. In short, the variability of climatic extremes in different regions tends to be different because of the huge differences in climate driving forces and regional terrain characteristics.
China is the third largest land area in the world. The complicated nature and geography of the environment causes many climatic disasters, and these are directly influenced by the spatiotemporal variation of precipitation [20
]. Great attention has been paid to extreme precipitation and temperature due to the global warming and rapid urbanization in China. The Yellow River is the second longest river in China and the fifth largest river in the world; characterized by uneven temporal-spatial patterns of precipitation, it has attracted more studies that focused on climate variations. Many researchers in the early 1960s focused on the entirety or sub-regions of the Yellow River basin (YRB) and investigated the variations of precipitation and/or temperature. Zhao et al. [28
] pointed out the annual temperature rose 0.80 °C in the upstream areas of the YRB from 1960 to 2001, but annual precipitation was not significant during the same period. Liu et al. [26
] examined the spatiotemporal variability of annual precipitation all along the YRB, and results showed that most regions exhibited prominent decreasing trends of precipitation in the whole basin from 1961 to 2006. Yang and Liu [29
] used 80 meteorological stations to determine that 83.75% of stations displayed decreasing trends with a reduction of 10.37% in annual precipitation during 1961–2000. Nevertheless, the precipitation extreme indices have a declining trend and the temperature extreme indices show the patterns of variability consistent with a general warming trend in the YRB over the past few decades [30
]. Meanwhile, Wang et al. [15
] noted decreasing trends in most extreme rainfall indices and increasing trends in extreme temperature indices in winter due to the increasing frequency of warm days and warm nights, and the decreasing frequency of cold days and cold nights. Dong et al. [31
] reported that extreme rainfall events showed a negative trend along the middle and lower reaches of the YRB from 1951 to 2004. Gao et al. [4
] explored the spatial and temporal variations and causes of precipitation extremes over the YRB from 1960 to 2011.
One of the physical mechanisms of climatic extremes is atmospheric circulation [32
], and lead to a variety of precipitation extremes [34
]. Some examples of atmospheric patterns are the El Niño–Southern Oscillation (ENSO), the Arctic Oscillation (AO), the Pacific Decadal Oscillation (PDO), the Indian Ocean Dipole (IOD), and the North Atlantic Oscillation (NAO) [9
]. These atmospheric circulation patterns (ACPs) are identified by the continuous, recurring, and large-scale modes of pressure anomalies. The ACPs are commonly expressed by numerical indices determining the intensity of the influence of atmospheric circulation during a specific period in a particular geographical region. Studies around the world revealed the influence of the large-scale atmospheric circulation changes on weather regimes, climate change, and hydrological variations [38
]. China climate types are mainly governed by the East Asian Monsoon. Nevertheless, the East Asian Summer Monsoon is significantly influenced by the ENSO [38
], the PDO [45
], the AO [46
], the IOD, and the NAO [43
]. Global ENSO events, which directly affected the regional precipitation in the river basin, resulted in an approximately 51% decrease in river water discharge to the sea [47
Here, we built on the existing literature to identify probable responses to the behavior of the atmospheric circulation patterns. Consequently, in this study, we extended these ACPs and attempted to solve the following three problems: (i) What are the trends in the changes of magnitude and frequency of temperature and precipitation extremes in the YRB? (ii) Can the ACPs that influence the climate of the YRB also have significant effects on extreme climate events in the YRB? If so, (iii) how and to what extent can they impact extreme climate events? Our principal objectives were to (i) analyze the temporal and spatial variations of daily temperature and precipitation extremes in the YRB during 1960–2017 and (ii) investigate the ACPs (ENSO, AO, NAO, PDO, and IOD) associated with these extreme indices by using the Pearson correlation analysis method and to interpret possible mechanisms based on the ACPs for observed changes in the YRB from 1960 to 2017. This research can provide a scientific basis for the studies of natural disasters in the whole basin of the Yellow River or its regions as a response to global climate change.
In the present study, a total of 16 indices that represent climatic extreme events were calculated by using daily precipitation and maximum and minimum temperature data from 66 stations from 1960 to 2017 and analyzed thoroughly for characteristic variation trends and their connections to the ACPs (ENSO, AO, NAO, PDO and IOD) over the entire YRB.
Our study concluded that there were no significant change trends in precipitation extremes during 1960–2017 in this basin. It should be noted that the significant decreasing trend (−2.88 days/decade) of CDD was observed in the upper regions (Table 2
). In addition, both the SDII and heavy precipitation events (e.g., R10mm and R25mm) had slightly increasing trends over the investigation period. Recently, precipitation changes were associated with latitude and elevation in some studies that found increased precipitation frequency in higher altitudes and a decreasing trend at lower altitudes in the southwest of China [53
]. The Intergovernmental Panel on Climate Change (IPCC) [30
] also pointed out that increases in the amount of precipitation were very likely at high latitudes while declines were likely in most subtropical land regions. These changes are consistent with previous studies in other parts of China [4
In this study, warming trends were observed in the indices of TXx and TNn, but the trend magnitude of TNn (0.45 °C/decade) was much larger than that of TXx (0.18 °C/decade). This asymmetry between the magnitudes of changes in TNn and TXx agrees with an earlier global study [55
] and a regional study [58
]. The trend magnitude of TNn in the YRB is lower than the average value of China and smaller than that on the global scale (0.71 °C/decade) [55
], but larger than that in the Yangtze River Basin (0.42 °C/decade) [18
], which is the longest river in China. In addition, we found that the occurrence of TX10p and TN10p significantly decreased, while the occurrence of TX90p and TN90p exhibited statistically significant increases during 1960–2017. These results were generally consistent with Li et al. [56
] and Guan et al. [18
]. Our results also revealed that the spatial pattern of changes in frequency and magnitude for climatic extremes presents obvious spatial differences. These results were in line with previous studies, for example, Tian et al. [59
] found that the annual precipitation did not show a significant trend and the annual mean temperature had a significant increasing trend over three major river basins of China (Yellow River, Yangtze River, and Pearl River) in the last few decades. Taken together, the increasing trends of extreme temperature indices (SU, TXx, TNn, TX90p, and TN90p) and decreasing trends of others (FD, TX10p, and TN10p) obtained in this study illustrate that there was more drought in the YRB in the last six decades despite a slight increase in the magnitude and frequency of heavy precipitation events.
Finally, we found that ENSO, AO, NAO, and PDO indices exhibited significant positive correlation for the TNn, and significant negative correlation for TN10p only at a few stations. ENSO indices exhibited significant positive correlation for the TX90p in most stations of the whole YRB, but AO and IOD indices showed significant negative correlation for TX90p in the lower reach of the YRB (Figure 4
, Figure 5
, Figure 6
and Figure 7
). Additionally, we found that those ACP indices exhibited positive correlation for the increasing trends of most extreme temperature indices, but negative correlation for those that showed decreasing trends. We conclude that the atmospheric circulation patterns have enhanced the gap of the maximum and minimum temperature and were the physical mechanisms for the extreme cold/warm events in the past decades. These results are consistent with previous studies in most parts of East Asia [38
], including the YRB of China [4
Based on the above analysis and discussions, the important conclusions can be summarized as follows.
For all the precipitation extreme indices, only a few stations were characterized by significantly increasing or decreasing extreme precipitation anomalies during 1960–2017 in the YRB. Nonsignificant increasing trends were detected for the most extreme indices including heavy precipitation and heaviest precipitation days, very wet and extremely wet day precipitation, and simple precipitation intensity index. However, nonsignificant decreasing trends were maximum 1-day and 5-day precipitation, and the consecutive dry days (e.g., CDD) exhibited a prominent significant decreasing trend (−2.88 days/decade) in the upper regions of the YRB. In addition, both precipitation intensity (e.g., SDII) and heavy precipitation magnitudes and frequencies (e.g., R95p, R10mm, and R25mm) increased slightly over the investigation period. These results reveal that the precipitation intensity may have been strengthened, and the duration of extreme rainfall events appears to have been reduced. Therefore, the heavy rainfall events tended to be increasingly shorter during 1960–2017 in the YRB.
All the temperature extreme indices revealed widespread significant changes associated with global warming in most regions of the YRB during the study period. For TNn, TN10p, TN90p, and FD, the warming trends were of greater magnitude than those of the indices TXx, TX10p, and TX90. For the vast majority of stations, significant increases in TX90p, TN90p, and SU, but significant decreases in TN10p, FD, and TX10p were observed over the entire study period.
In this study, we found that the relationship between the ACPs (ENSO, AO, NAO, PDO, and IOD) and precipitation extremes showed that the ACPs are not clearly associated with precipitation extremes. However, the relationship between the ACPs and temperature extremes shows that the influence of the ENSO and AO on temperature extremes outweighs that of the NAO, IOD, and PDO for most extreme indices in the YRB. The ENSO, AO, NAO, and PDO indices exhibited significant positive correlation for TNn, but significant negative correlation for TN10p. Therefore, the relationship between the ACPs and climatic extremes demonstrates that the influence of the ACPs on temperature extremes outweighs that on precipitation extremes in the YRB. We can conclude that the change of the ACPs is an important physical mechanism affecting the heat and moisture transportation in this region. Admittedly, the changes of climatic extremes are driven by the broader context of climate change; we did not make more detailed assessments of the physical mechanisms between the ACPs and climatic extremes in this paper, and further work needs to be carried out to clarify this issue.