Changes in temperature extremes

On a regional basis, a number of studies of extremes have been completed as a part of a series of regional workshops coordinated by the joint World Meteorological Organization (WMO) Climate Variability and Predictability/ Climate Change Indices (CLIVAR/CCI) Expert Team on Climate Change Detection, Monitoring and Indices (ETCCDMI). Each of these workshops used a set software that calculated extremes defined in Frich et al. (2002). These workshops included the Caribbean (Peterson et al., 2002), southern South America (Vincent et al., 2005), Central America and northern South America (Aguilar et al., 2005), central and northern Africa (Easterling et al., 2003), southern and western Africa (New et al., 2007), the Middle East (Zhang et al., 2005), Australasia and southeast Asia (Griffiths et al., 2005), and central and southern Asia (Klein Tank et al., 2006). In addition, other researchers have performed regional studies for North America (Vincent and Mekis, 2007), the Arctic (Groisman et al., 2003), Western Europe and East Asia (Kiktev et al., 2003), and China (Zhai and Pan, 2003). Results from these studies are consistent with the observed increases in global temperatures, showing evidence of changes in temperature and precipitation extremes defined by the 10th and 90th percentiles.

More specifically, in the United States, two studies focused on the northeastern United States support the notion that changes in the number of days exceeding certain thresholds have occurred. Cooter and LeDuc (1995) showed that the start of the frost-free season in the northeastern United States occurred 11 days earlier in the 1990s than in the 1950s. In an analysis of 22 stations in the northeastern United States for 1948-93, DeGaetano (1996) found significant trends to fewer extreme cold days, but trends to fewer warm maximum temperatures as well. More recently, Easterling (2002) found decreases in the number of days where the minimum temperature fell below freezing (0 °C) in the United States for the period 1948-99, with the largest decreases occurring in the western United States.

Results for the frost-free season showed increases in length for all areas of the continental United States, being driven mainly by changes to earlier dates of the last freeze in the spring season. Kunkel et al. (2004), using newly available data for the pre-1948 period, found that the average length of the frost-free season during 1895-2000 for the United States has increased by almost 2 weeks (Fig. 2.4). The change is characterized by four distinct regimes, with decreasing

Length Spring Frost -Fall Frost

Figure 2.4. Time series of frost-free season length; Julian date of the first fall frost and last spring frost for the conterminous United States. (From Kunkel et al, 2004.)

Length Spring Frost -Fall Frost

Figure 2.4. Time series of frost-free season length; Julian date of the first fall frost and last spring frost for the conterminous United States. (From Kunkel et al, 2004.)

frost-free season length from 1895 to 1910, an increase in length of about 1 week from 1910 to 1930, little change during 1930 to 1980, and strong increases since 1980. Both Easterling (2002) and Kunkel et al. (2003) found that the frost-free season length has increased more in the western United States than in the eastern United States, which is consistent with results of Cayan et al. (2001), who found the spring pulse of snowmelt water in the western United States now comes as much as 7-10 days earlier than in the 1950s. Nemani et al. (2001) showed that frost days in the Napa/Sonoma region of California diminished from 28 days in 1950 to about 8 days by 1997, together with a substantial increase (66 days) in the length of the continuously frost-free season - from 254 to 320 days per year-which has benefited the premium wine industry.

An analysis of multiday extreme heat and cold episodes when the temperature exceeded the 10-year return period did not show any overall trend for the period 1931-98 (Kunkel et al., 1999). The most notable feature of the temporal distribution of these very extreme heat waves is the high frequency in the 1930s compared with the rest of the record. DeGaetano and Allen (2002) examined daily exceedances of 90th, 95th, and 99th percentile thresholds (defined monthly) in the United States. The number of days exceeding these thresholds has increased in recent years, although they were also dominated earlier in the twentieth century by the extreme heat and drought of the 1930s. Changes in cold extremes (days below the 10th, 5th, and 1st percentile threshold temperatures) have shown decreases, particularly since 1960. Bonsal et al. (2001) used daily data adjusted for inhomogeneities to examine changes in temperature extremes in Canada. They found fewer cold extremes in winter, spring, and summer in southern Canada and more high temperature extremes in winter and spring, but little change in warm extremes in summer. Robeson (2004) examined changes in daily maximum and minimum temperatures by percentile for the United States and Canada and found that the largest increases in temperature have been occurring in the colder days of each month.

Globally, recent work by Alexander et al. (2006) showed that changes in the number of days (for night-time temperatures) exceeding selected temperature percentile thresholds corresponds well to areas with strong warming in minimum temperatures. In particular, the number of cold nights (less than the 10th percentile threshold temperature) has decreased for most of the areas examined, including the middle and higher latitudes of both the Northern and Southern Hemispheres, and the number of warm nights (nights warmer than the 90th percentile threshold) have increased in the same areas. There is less consistency with changes in the number of warm or cold days (daytime highs above the 90th percentile or below the 10th percentile); however, the global trends in both the night-time and daytime temperatures have changed in a similar fashion, with fewer cold days (nights) and more warm days (nights).

Was this article helpful?

0 0

Post a comment