Received 19 March 2016; accepted 15 August 2016; published 18 August 2016
At present, no doubt exists about global warming. The fourth assessment report of the Intergovernmental Panel on Climate Change  identified a warming trend of 0.13˚C/10a over the past 50 years. The warming trend in China has been 0.22˚C/10a  , with significant increases in annual mean temperature and heating rate  . The increasing mean temperature not only affects the change in extreme temperature values, but can also leads to extreme climatic events, such as heat waves and rainstorms, which show trends of increasing frequency, strength, and intensity  .
Many studies have indicated that frequent extreme climatic events cause huge losses for society and the economy, as well as loss of human lives  . The most recent statistics reveal an increase of up to 10-fold in the economic loss arising from global climatic change and related extreme climatic events over the past 40 years. In China, weather disasters caused by extreme climatic events account for 70% of all natural hazards  . Hence, extreme climatic events are sources of widespread concern for governments and researchers   . Studies of changes in these events are necessary not only to guarantee state security and economic development, but also to prevent disasters, reduce damage-related needs, and, most importantly, safeguard people’s safety and livelihoods.
Xinjiang, located far from the ocean in the center of Eurasia, in the border region of northwestern
2. Research Data and Methods
2.1. Research Data Source and Treatment
Daily precipitation and maximum, minimum, and mean temperature data from China’s Meteorological Administration for the period 1961-2010 in Xinjiang were used for this study. Data from 43 meteorological stations were reviewed to identify problems with missing or misdetected observations. We selected high-quality data from 35 stations (Table 1, Figure 1) as the study target. For missing and misdetected data from stations with less complete records, we used corrected data from Bai Lei et al.  .
2.2. Study Method
The standards used to define and calculate the extreme climatic indices used in this study were based on the World Meteorological Organization’s Commission for Climatology World Climate Research and Climatic Variation and Predictability programs’ expert team on climate change detection, monitoring, and indices  . This methodology has been applied to many countries and regions  , including Xinjiang. Using RClimDex software (http://cccma.seos.uvic.ca/ETCCDMI/), 19 extreme climatic indices (7 related to precipitation and 12 related to temperature) were developed and applied (Table 2).
All of the extreme climate indices reflect three aspects of temperature or precipitation events: strength, frequency, and temporal duration of climatic change. In extreme climate research, a certain percentile value is usually adopted as a threshold (with values exceeding it defined as extreme), and calculations are performed according to the nonparametric method of Bonsal et al.  , as follows: for a meteorological factor with N values, these values are sorted in ascending order (X1, X2, …, Xm, …, XN); the probability that a certain value is less than or equal to the rank of Xm is then P = (m − 0.31)/(N + 0.38), where P denotes percentile, m is the serial number of Xm, and N is the number of meteorological values.
In this study, percentile values were used to calculate relative and other indices. Specifically, temperature data were sorted in ascending order, and the 90th and 10th percentile values were regarded as thresholds for extreme temperature. When the highest temperature on a given day exceeded the 90th percentile value, an extreme high-temperature event was considered to have occurred on that day; when the highest temperature on a given day was less than the 10th percentile value, an extreme low-temperature event was considered to have occurred. Secondly, we analyzed extreme climatic events using linear tendency estimation, the Mann-Kendall method, and the kriging method of spatial analysis.
Table 1. Site meteorological information in the study area.
Figure 1. Distribution of meteorological stations in Xinjiang.
Table 2. Definition of extreme temperature and precipitation indices.
Notes: aAll the indices are calculated by RCLimDEX. Abbreviations are as follows: TX, daily maximum temperature; TN, daily minimum temperature; TG, daily mean temperature; RR, daily precipitation. A wet day is defined when RR ≥ 0.1 mm, and a dry day is defined when RR < 0.1 mm. Indices are included for completeness but are not analyzed further in this article.
3. Changes in Extreme Temperature Events in Xinjiang over the Past 50 Years
3.1. Interannual Variation in Extreme Cold Events
Consistent with global changes, the frequency of extreme low-temperature events in the Xinjiang area was found to have decreased while that of extreme high-temperature events increased  . Linear variation in extreme cold indices in Xinjiang over the past 50 years generally showed a decreasing curvilinear trend (Figure 2). The interannual variation tendency rates for the maximum numbers of continuous frost days (FDs), icy days (IDs), cold-nighttime days (TN10p), and cold-daytime days (TX10p) were −4.8 d/10a, −2.75 d/10a, −2.24 d/10a, and −0.88 d/10a, respectively (all α ≤ 0.01), and their correlation coefficients by year were −0.774, −0.475, −0.827, and −0.39, respectively (Table 3).
(Note: the straight line means linear fitting trend line, the smooth curve is moving average value in 5 years, attached mark is actual test value)
Figure 2. Regional annual anomalies series during 1961-2010 for indices of cold extremes.
Table 3. The extreme temperature mean value (˚C or d) and change range [˚C/
The monthly maximum temperature minimum value (TXn) and monthly extreme minimum temperature (TNn) showed increasing trends, with interannual variation tendency rates of 0.16˚C/
3.2. Spatial Variation in Extreme Cold Events
Against the background of global warming, the extreme cold indices generally showed decreasing trends over the past 50 years, but changes in these indices were characterized by spatial variation (Figure 3). Change tendency rates of continuous FD ranged from −7.8 to −2.0 d/
Variation tendency rates for TN10p ranged from −3.7 to −0 d/
3.3. Interannual Variation in Extreme Warm Events
For the past 50 years, all extreme warm event indices, such as the numbers of summertime days (SU25), warm- nighttime days (TN90p), and warm-daytime days (TX90p), as well as monthly extreme maximum temperature (TXx) and the monthly minimum temperature maximal value (TNx), showed significant increasing trends in the Xinjiang area (Figure 4). The increasing rates for SU25, TN90p, and TX90p were 2.59, 3.38, and 1.59 d/10a, respectively (all α ≤ 0.05), with correlation coefficients by year of 0.48, 0.86, and 0.6, respectively (all with ≥99% confidence). The variation tendency rates for TXx and TNx were 0.23˚C/10a and 0.52˚C/10a, respectively, and the correlation coefficients by year were 0.36 and 0.64, respectively (all α ≤ 0.01).
The graph of changes in day-by-day temperature range (DTR; Figure 3 & Figure 4) shows a decreasing trend in Xinjiang over the past 50 years (rate, 0.21˚C/10a), with a correlation coefficient by year of −0.54 (α = 0.01). The trend of change in extreme annual temperature range in the entire Xinjiang area complies with that throughout China  .
3.4. Spatial Variation in Extreme Warm Events
Against the background of global warming, the extreme warm indices in the Xinjiang area generally showed increasing trends over the past 50 years, but these changes were characterized by spatial variation (Figure 5). Variation tendency rates for SU25 ranged from0 to 8.0 d/
Although the ranges of increase in variation tendency rates for TXn and TNn exceeded those for TXx and TNx, all stations showed decreasing trends for DTR.
Figure 3. Spatial distribution of linear tendency rate for indices of cold extremes during 1961-2010.
(Note: the straight line means linear fitting trend line, the smooth curve is moving average value in 5 years, attached mark is actual test value)
Figure 4. Regional annual anomalies series during 1961-2010 for indices of warm extremes.
4. Changes in Extreme Precipitation Events in Xinjiang over the Past 50 Years
4.1. Interannual Variation Trends for Extreme Precipitation Indices
Precipitation intensity is one factor used to measure extreme precipitation, with greater intensity associated with greater possibility of disaster. Average precipitation intensity in the Xinjiang area showed a predominant trend of annual increases over the past 50 years, consistent with global data and those for China  . However, the range of change indicated by the extreme precipitation intensity indices was small relative to those for extreme temperature events. Total precipitation during the wet season (PRCPTOT) in each year from 1961 to 2010 appeared to show a significant increasing trend (rate,
Figure 5. Spatial distribution of linear tendency rate for indices of warm extremes during 1961-2010.
Note: the red dot means actual measure value, green line means moving average in 5 years, the straight line means simple regression trend line.
Figure 6. Inter-annual variation of precipitation extremes in Xinjiang during 1961-2010.
Table 4. The standard difference of extreme precipitation index and change tendency rate (mm/10a) and mean value (d or mm) in Xinjiang.
The simple precipitation intensity index (SDII) showed small increasing trend at a rate of
4.2. Spatial Distribution Pattern of Extreme Precipitation Indices
Interannual variation tendencies for most extreme precipitation indices, with the exception of the CDD index, showed significant increasing trends at most stations in the Xinjiang area over the past 50 years (Figure 7). PRCPTOT showed decreasing trends at only 2.8% of the 35 stations from which data were analyzed; most stations showed increasing trends (range, −30.68 to
Increasing trends in the extreme precipitation indices of RX1day, RX5day, R95, and CWDs were found at 25 (71.4%), 28 (80%), 30 (85.7%), and 29 (82.8%) stations, respectively. More than 70% of stations showed increasing trends for PRCPTOT, SDII, RX1day, RX5day, R95, and CWDs, indicating that the frequency of extreme precipitation trends is generally increasing in the Xinjiang area. Analysis of CDD data from 11 (31.4%) stations in southern and northern Xinjiang revealed an increasing trend and demonstrated regional tendencies for increased numbers of disasters involving drought.
In this study, 12 extreme temperature indices and 5 extreme precipitation indices were used to examine temporal and spatial variation in climatic extremes in Xinjiang over the past 50 years. The main conclusions are described below.
1) Analysis of temporal changes revealed decreasing trends in extreme cold indices (i.e., obvious reductions in the number of severely cold days and extreme low-temperature events), consistent with global warming. Linear variation in extreme warm indices showed notable increasing trends. The trends for TX10p, TN10p, IDs, and FDs (−0.88, −2.24, −2.75, −4.8 d/10a, respectively) decreased, whereas those for TX90p, TN90p, and SU25 (1.59, 3.38, and 2.59 d/10a, respectively) increased.
2) Analysis of temporal changes in extreme precipitation indices (except CDDs), such as RX1day, RX5day, R95, and CWD, as well as PRCPTOT, showed consistently increasing trends (rates of 0.474 mm/10a, 0.434 mm/10a, 9.322 d/10a, 0.154 d/10a, and 8.232 mm/10a, respectively).
3) Differences in the spatial distributions of the indices were notable. The frequencies of extreme cold and extreme warm events decreased in southern Xinjiang. The spatial distribution of extreme precipitation also showed obvious regional differences, with the directionality of trends differing between mountainous and desert basin areas. The response to global warming has been more notable in northern than in southern Xinjiang. Global climatic change has altered the ecology of northern Xinjiang and increased the frequency of extreme climatic disasters.
4) Changes in extreme cold and warm indices, as well as those in nighttime and daytime indices, showed notable asymmetry: the warming ranges of the cold and nighttime indices exceeded those of the warm and daytime indices, respectively.
Figure 7. Spatial distribution of inter-annual variation of precipitation extremes in Xinjiang during 1961-2010.
This study is jointly financed by the National Natural Science Fund Project (U1203282) and the National Natural Science Foundation Project of China (41001020), and the Xinjiang Uygur Autonomous Region Key Laboratory of “Xinjiang laboratory of Lake Environoments and Resources in Arid Zone” (XJDX0909-2012- 12), the Shihezi University team innovation project (2014ZRKXJQ08). The authors gratefully acknowledge funding for this research and would like to express their sincere thanks to Zhang Yan Wei for the help with data.
 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. In: Solomon, S., Qin, D., Manninget, M., et al., Eds., Cambridge University Press, Cambridge, New York.
 Li, Z.X., He, Y.Q., Wang, C.F., et al. (2011) Spatial and Temporal Trends of Temperature and Precipitation during 1960-2008 at the Hengduan Mountains, China. Quaternary International, 236, 127-142.
 Easterling, D.R., Gerald, A.M., Camille, P., et al. (2000) Climate Extremes: Observations, Modeling, and Impacts. Science, 289, 2068-2074.
 Jiang, F.Q., Hu, R.J., Zhang, Y.W. and Li, X.M. (2011) Variations and Trends of Onset, Cessation and Length of Climatic Growing Season over Xinjiang, NW China. Theoretical and Applied Climatology, 106, 449-458.
 Aguilar, E., Peterson, T.C., Ramirez Obando, P., et al. (2005) Changes in Precipitation and Temperature Extremes in Central America and Northern South America, 1961-2003. Journal of Geo-Physical Research, 110, D23107.
 Zhang, Y.W., Wei, W.S., Jiang, F.Q., et al. (2012) Assessment of Change in Temperature and Precipitation over Xinjiang, China. Natural Hazards and Earth System Sciences, 12, 1327-1331.
 Alexander, L.V., Zhang, X., Peterson, T.C., et al. (2006) Global Observed Changes in Daily Climate Extremes of Temperature and Precipitation. Journal of Geophysical Research, 111, DO5109.
 Zhang, Z.Y., Liu, L. and Tang, X.L. (2013) Tianshan Mountains in China Nearly 50 Years Precipitation Change Regional Differences and Mutation Characteristics. Journal of Resources and Environment in Arid Areas, 7, 85-90.