Longterm changes in rainfalls

There is much speculation about the effects of climate change on rainfall amounts and frequencies. Such changes need to be put into the context of an analysis of historical variability, because we know there has been a lot of variability in rainfalls in the past. Detection of possible changes over time in short duration rainfall extremes is difficult because of the variability (see Section 9.6 on frequency analysis). Analysis of longer term totals can sometimes suggest that there has been change, but it has not all been recent change. The first analyses of this type date back to well before climate change was a real issue (e.g. the study by Nicholas and Glasspoole, 1931).

Fig. 9.16 shows a more recent analysis of seasonal rainfall totals in Scotland taken from Jones and Conway (1997). There appears to be a distinct upward trend in rainfall

700 600 500 | 400 300 200 100

700 600 500 | 400 300 200 100

700 600 500 | 400 300 200 100

700 600 500 | 400 300 200 100

Fig. 9.16 Changes in 3-monthly rainfall totals over Scotland, 1760-1995. The smooth line represents a decadal Gaussian filter centred on each year of data (after Jones and Conway, 1997, with kind permission of John Wiley & Sons).

i i i i i ■ ■ p i i i ■ i ■ i i i i i ■ r

Winter (DJF)

i i i i 1 i i i i i i i i i i i i i i i i ' i i i i i i ■ ■ p i i i ■ i ■ i i i i i ■ r

Winter (DJF)

1760 1780 1800 1820 1840 1860 1880 1900 1920 1940 1960 1980

l i i i i i i r r i i i l l i i r r l l i i i i l i i i i i r i i r i l l i i i i r r i i r r l

Spring (MAM)

1760 1780 1800 1820 1840 1860 1880 1900 1920 1940 1960 1980

l i i i i i i r r i i i l l i i r r l l i i i i l i i i i i r i i r i l l i i i i r r i i r r l

Spring (MAM)

1760 1780 1800 1820 1840 1860 1880 1900 1920 1940 1960 1980

1760 1780 1800 1820 1840 1860 1880 1900 1920 1940 1960 1980

700 600 500 | 400 300 200 100

1760 1780 1800 1820 1840 1860 1880 1900 1920 1940 1960 1980

Autumn (SON)

i i i i i r r i i iiiiii^ii i » f i i i i i i i i * p p t p t i p i

Autumn (SON)

700 600 500 | 400 300 200 100

Fig. 9.16 Changes in 3-monthly rainfall totals over Scotland, 1760-1995. The smooth line represents a decadal Gaussian filter centred on each year of data (after Jones and Conway, 1997, with kind permission of John Wiley & Sons).

1760 1780 1800 1820 1840 1860 1880 1900 1920 1940 1960 1980

1760 1780 1800 1820 1840 1860 1880 1900 1920 1940 1960 1980

in the winter months and a downward trend in the summer months, but with a high degree of variability from year to year. In both cases, an analysis for a purely linear trend suggests that the changes are statistically significant. Other regions of the UK show different seasonal patterns of change, not always with significant trends (such as the analysis of data since 1838 at Armagh Observatory in Butler et al., 1998) and with complicating factors such as rain shadow effects (Malby et al., 2007). There is no reason, of course, why we should expect any trends in the data to be linear and there are also other methods for analysis of trends in data that do not assume a purely linear change over time, such as the dynamic harmonic regression method of Young et al. (1999) and the general linear models of Chandler and Wheater (2002).

0 0

Post a comment