Precipitation, temperature and runoff analysis from 1950 ...precipitation leads to increasing cloud coverage and, hence, results in decreasing ground surface temperature. Average monthly

Hydrological Sciences–Journal–des Sciences Hydrologiques, 50(1) February 2005

Open for discussion until 1 August 2005 Copyright 2005 IAHS Press

65

Precipitation, temperature and runoff analysis from 1950 to 2002 in the Yangtze basin, China QIANG ZHANG1, TONG JIANG1, MARCO GEMMER2 & STEFAN BECKER3

1 Nanjing Institute of Geography and Limnology, Chinese Academy of Sciences, Nanjing 210008, China [email protected]

2 Institute for Geography, Justus-Liebig-University, Senckenbergstrasse 1, D-35390 Giessen, Germany

3 Department of Geography & Urban Planning, University of Wisconsin-Oshkosh, Oshkosh, Wisconsin 54901, USA

Abstract Monthly precipitation and temperature trends of 51 stations in the Yangtze basin from 1950–2002 were analysed and interpolated. The Mann-Kendall trend test was applied to examine the monthly precipitation and temperature data. Significant positive and negative trends at the 90, 95 and 99% significance levels were detected. The monthly mean temperature, precipitation, summer precipitation and monthly mean runoff at Yichang, Hankou and Datong stations were analysed. The results indicate that spatial distribution of precipitation and temperature trends is different. The middle and lower Yangtze basin is dominated by upward precipitation trend but by somewhat downward temperature trend; while downward precipitation trend and upward temperature trend occur in the upper Yangtze basin. This is because increasing precipitation leads to increasing cloud coverage and, hence, results in decreasing ground surface temperature. Average monthly precipitation and temperature analysis for the upper, middle and lower Yangtze basin, respectively, further corroborate this viewpoint. Analysis of precipitation trend for these three regions and of runoff trends for the Yichang, Hankou and Datong stations indicated that runoff trends respond well to the precipitation trends. Historical flood trend analysis also shows that floods in the middle and lower Yangtze basin are in upward trend. The above findings indicate that the middle and lower Yangtze basin is likely to face more serious flood disasters. The research results help in further understanding the influence of climatic changes on floods in the Yangtze basin, providing scientific background for the flood control activities in large catchments in Asia. Key words floods; Mann-Kendall analysis; monthly-mean flood runoff; precipitation; temperature; trend analysis; Yangtze catchment

Analyse des précipitations, températures et débits de 1950 à 2002 dans le bassin du Yangtze, en Chine Résumé Les tendances des précipitations et des températures mensuelles de 51 stations du bassin du Yangtze, entre 1950 et 2002, ont été analysées et interpolées. Le test de variation de Mann-Kendall a été appliqué pour examiner les données mensuelles de précipitation et de température. Des tendances positives et négatives significatives ont été détectées aux niveaux de significativité de 90, 95 et 99%. Les températures, précipitations et précipitations estivales moyennes mensuelles, ainsi que les débits moyens mensuels des stations de Yichang, Hankou et Datong, ont été analysés. Les résultats montrent que les distributions spatiales des tendances des températures et des précipitations sont différentes. Les deux zones moyenne et aval du bassin du Yangtze sont dominées par une tendance à la hausse des précipitations mais par une certaine tendance à la baisse des températures; tandis que la partie amont du bassin du Yangtze est marquée par des tendances à la baisse pour les précipitations et à la hausse pour les températures. Cela vient du fait que l'augmentation des précipitations induit l'augmentation de la couverture nuageuse et, par conséquent, la diminution de la température à la surface du sol. L’analyse des moyennes mensuelles des précipitations et des températures, respectivement dans les zones amont, moyenne et aval du bassin, corrobore plus encore ce constat. L’analyse de la tendance des précipitations pour ces trois régions et des tendances des débits des stations de Yichang, Hankou et Datong montre que les variations de débits répondent bien aux variations pluviométriques. L’analyse de l’historique des crues indique également une

Qiang Zhang et al.

Copyright 2005 IAHS Press

66

tendance à la hausse des épisodes catastrophiques sur les parties moyenne et aval du bassin du Yangtze. Les résultats de ces recherches aident à mieux comprendre l’influence des changements climatiques sur les crues dans le bassin du Yangtze, tout en fournissant un support scientifique aux activités de contrôle des crues dans les grands bassins asiatiques. Mots clefs crues; analyse de Mann-Kendall; débit de crue mensuel moyen; précipitation; température; analyse de tendance; bassin du Yangtze

INTRODUCTION Floods are among the most serious natural disasters and have a severe impact on human society worldwide. According to Zhou et al. (2002) and CWRC (2002), of 594 recorded severe natural disasters in the world in 1996, 27% were floods, causing 56% of the total number of deaths and 57% of the total economic losses. The occurrence of floods has been studied from various points of view (e.g. Prudhomme et al., 2003; Grossman, 2001; Simonovic & Carson, 2003). In particular, the connections between climatic changes and occurrence of floods are receiving increasing attention in academic circles (e.g. Muzik, 2002; Boorman & Sefton, 1997). The Yangtze (Changjiang) River originates in the Qinghai–Tibet Plateau and flows about 6300 km eastwards to the East China Sea. The drainage basin lies between 91–122°E and 25–35°N (Fig. 1), covering 1 808 500 km2 with mean discharge of 23 400 m3 s-1 at Hankou station (at Wuhan, on the middle Yangtze River). The Yangtze basin is climatically characterized by a subtropical monsoon climate. The southern part of the basin is climatically close to tropical climate and northern part is near to temperate zone. The annual mean temperature in the southern and northern

Fig. 1 Location of the study region—Yangtze basin (shaded zone).

Precipitation, temperature and runoff analysis from 1950 to 2002 in the Yangtze basin, China


67

parts of the middle and lower Yangtze basin is 19 and 15°C, respectively. The topo-graphy in the upper Yangtze basin exerts great influence on the climate, e.g. the annual mean temperature is in the range from 17 to 0°C between the Sichuan basin and the Tibet Plateau. The long-term mean annual precipitation of the whole Yangtze basin is 1067 mm, but the spatial distribution of the rainfall is extremely uneven. Generally speaking, the rainfall decreases from the southeast to the northwest of the basin. Since differences in longitude and altitude (less in latitude) give rise to sharp variations in precipitation and temperature within the region, the whole Yangtze basin is divided into three regions for the purposes of this study, which corresponds to the altitude zones (Table 1). Temporal and spatial distribution of the rainstorm zone is in close relation with monsoon activities and the seasonal motion of subtropical highs. Therefore, summer is the main flooding season for the Yangtze basin. The Yangtze floods caused serious problems in China. In 1998, the entire Yangtze drainage area suffered from tremendous flooding—the largest flood since 1954, leading to the economic loss of 166 billion (×109) yuan (US$20 billion) (Yin & Li, 2001). In this study, the instrumental data of precipitation, temperature and runoff, and historical flood data are collected and analysed to better understand the impacts of climatic change on flood occurrence. DATA AND METHODS A data set of 51 National Meteorological Observatory (NMO) stations with long-term monthly precipitation and temperature data (1950–2002) in the Yangtze basin, China, were analysed for this study (Fig. 2, Tables 1 and 2). The data were provided by the National Climatic Centre of China of the China Meteorological Administration and contain values from January 1950 to December 2002. The location of the stations can be seen in Fig. 2. The homogeneity of the precipitation records was analysed by calcu-lating the von Neumann ratio (N) and the cumulative deviations (Q/n-0.5 and R/n-0.5) (Buishand, 1982). Monthly mean flood runoff data of 1950–2000 were also collected from three main gauging stations in the Yangtze basin: Yichang Station (coverage area: 1 005 501 km2), Hankou station (1 488 036 km2) and Datong station (1 705 383 km2) (Fig. 2). These data were provided by the Changjiang Water Resources Commission (CWRC). Historical data of floods were extracted from documents (Luo & Le, 1996) and a modern flooding survey (CWRC, 2000). The main analysis method used in this paper is the Mann-Kendall rank statistic method (MK) (Libiseller, 2002; Kendall, 1938). In the 1970s, the catastrophe analysis of climatic changes was widely used in the research on climatic evolution (Helsel & Hirsch, 1992; Wei, 1999). However, in this study, the MK method has been used to explore the changing features of precipitation and temperature in June, July and August in the upper, middle and lower Yangtze basin, as well as the summer (June, July, August) precipitation and flood runoff from Yichang, Hankou and Datong stations (Fig. 2). The MK method assumes that the time series under research are stable, independent and random with equal probability distribution. Assume the time series

Qiang Zhang et al.


68

1 Ankang 14 Guangchan 27 Liling 40 Xichang 2 Anqing 15 Guixi 28 Lushan 41 Xinyang 3 Bijie 16 Guiyang 29 Mianyang 42 Ya’an 4 Changde 17 Hangzhou 30 Minxian 43 Yibin 5 Changsha 18 Hankou 31 Nanchang 44 Yichang 6 Chengdu 19 Hanzhong 32 Nanchong 45 Youyang 7 Chongqing 20 Hefei 33 Nanjing 46 Yueyang 8 Dali 21 Hengyang 34 Nanyang 47 Yunxian 9 Daxian 22 Huili 35 Neijiang 48 Yushu 10 Deqen 23 Ji'an 36 Panxian 49 Zhijiang 11 En'shi 24 Kangding 37 Rongjiang 50 Zhongxian 12 Ganzhou 25 Kunming 38 Shanghai 13 Ganzi 26 Lijiang 39 Tianshui

Fig. 2 Stations in the Yangtze River basin (numbered solid dots—see key above; solid squares denote Yichang, Hankou and Datong hydraulic stations).

under study is x1, x2, x3, …, xn, and mi denotes the cumulative total of samples so that xi > xj (1 ≤ j ≤ i), where n is the number of the samples. The definition of the statistical parameter dk is as follows:

�=k

iik md (2 ≤ k ≤ N) (1)

on the condition that the original time series was random and independent. The variance and mean of dk are defined as:

4)1(][E −= kkdk (2)

and

72)52)(1(]var[ +−= kkkdk (2 ≤ k ≤ N) (3)

Under the assumption above, the definition of statistical index of UFk is as follows:

]var[][

k

kkk d

dEdUF −= (k = 1, 2, 3, …, n) (4)



69

Table 1 Three regions of the Yangtze basin analysed in the paper.

West Middle East Longitude ~104°E 104–113°E ~113°E Mean altitude (m a.s.l.) 2551 627 113 Number of stations 10 22 19 Table 2 Meteorological stations analysed in the paper with monthly precipitation, temperature (1950–2002).

Station name Latitude (°N) Longitude (°E) Elevation (m a.s.l.) Ankang 32.53 109.35 290.8 Anqing 30.52 117.03 19.8 Bijie 27.30 105.23 1511.0 Changde 28.92 111.55 35.0 Changsha 28.25 112.83 44.9 Chengdu 30.67 104.07 506.1 Chongqing 29.78 106.33 259.1 Dali 25.72 100.18 1190.5 Daxian 31.27 107.47 168.7 Deqen 28.82 98.82 3592.9 En’shi 30.27 109.37 457.1 Ganzhou 25.83 114.83 123.8 Ganzi 31.63 99.98 3393.5 Guangchan 26.80 116.18 143.8 Guixi 28.28 117.10 49.0 Guiyang 26.57 106.70 1071.2 Hangzhou 30.32 120.20 41.7 Hankou 30.63 114.28 23.3 Hanzhong 33.00 107.27 508.4 Hefei 31.88 117.25 27.9 Hengyang 26.93 112.50 103.2 Huili 26.68 102.25 1787.1 Ji’an 27.08 114.92 76.4 Kangding 30.08 102.03 2617.0 Kunming 25.02 102.68 1891.4 Lijiang 26.95 100.30 2393.2 Liling 26.23 111.60 175.0 Lushan 29.58 115.98 1165.0 Mianyang 31.48 104.72 470.8 Minxian 34.38 104.17 2314.6 Nanchang 28.58 115.88 46.7 Nanchong 30.80 106.08 297.7 Nanjing 32.00 118.80 8.9 Nanyang 33.03 112.55 129.2 Neijiang 29.58 105.05 347.1 Panxian 25.75 112.98 1527.1 Rongjiang 25.87 108.58 327.0 Shanghai 31.17 121.43 4.5 Tianshui 34.58 105.75 1131.7 Xichang 27.90 102.27 1590.0 Xinyang 32.12 114.08 114.5 Ya'an 30.00 103.05 627.6 Yibin 28.82 104.53 340.8 Yichang 30.70 111.08 133.1 Youyang 28.80 108.77 663.7 Yueyang 29.38 113.08 51.6 Yunxian 32.82 110.77 201.9 Yushu 33.10 96.75 3681.2 Zhijiang 27.45 109.63 272.2 Zhongxian 31.17 112.57 65.8 Zunyi 27.68 106.92 843.9

Qiang Zhang et al.


70

Here the term UFk satisfies the normal distribution. The observed trends were spatially interpolated by applying the inverse distance weighted (IDW) interpolation method after applying and comparing different interpolation methods (e.g. kriging). Applying other interpolation methods results in similar output maps (Gemmer et al., 2004). RESULTS Observed precipitation trends The observed trends were spatially interpolated by applying the IDW interpolation method using the ArcView software package. The changing trends were tested at 90, 95 and 99% significance levels. It can be seen from Fig. 3 that in June the precipitation in

Fig. 3 Spatial distribution of Mann-Kendall changing trends of precipitation in (a) June, (b) July and (c) August at 90, 95 and 99% significance levels.

(a)

(b)

(c)



71

the middle Yangtze basin shows a significant upward trend (at >90% significance level). In some stations in the southern part of the middle Yangtze basin, this upward trend can reach 99% significance level. Precipitation in some parts of the lower Yangtze basin also shows an upward trend at >95% significance level. In other places, however, the precipitation trend is not significant. However, in July, the places with upward precipitation trend are shifted to the lower Yangtze basin as compared to June. Upward precipitation trend in some stations of the lower Yangtze, such as Shanghai, reached the 99% significance level. Some places in the upper Yangtze basin showed a downward trend at >95% significance level. In August, places with upward precipita-tion trend were concentrated mainly in the Yangtze Delta region at >95% significance level. The precipitation trend in other places in August was not statistically significant (at <90% significance level). Figure 4 demonstrates the spatial distribution of the changing trend of summer pre-cipitation (Fig. 4(a)) and summer precipitation anomalies between the 1990s and 1951–1990 (Fig. 4(b)). One can see that, in the middle and lower Yangtze basin, the summer precipitation has an upward trend at >90% significance level. The downward trends occurred in the upper Yangtze basin, especially in the northern Yangtze basin at >90% significance level. With regard to the summer precipitation anomalies, one can see that middle and lower Yangtze basin are characterized by increasing summer precipitation (with precipitation anomalies of <51–100 mm and even >101 mm) and

Fig. 4 Spatial distribution of Mann-Kendall changing trends of (a) summer precipita-tion (June, July, August) at 90, 95 and 99% significance levels and (b) summer precipitation (June, July, August) anomalies between the 1990s and 1951–1990.

(a)

(b)

Qiang Zhang et al.


72

1950 1960 1970 1980 1990 200090

120150180210

June

Y=365.57-0.01XR=-0.06

Year

120150180210240270

July

Y=-46.6+0.12XR=0.06

70140210280350

Aug

ust

Y=394.2-0.1XR=-0.04

500600700800900

tota

l sum

mer

prec

ipita

tion Y=771.3-0.04X

R=-0.009

A

1950 1960 1970 1980 1990 200080

120160200240

June

Y=-721.5+0.44XR=0.23 80

160

240

July

Year

Y=185.26+0.01XR=-0.003

70140210

Aug

ust

Y=1110.96-0.49XR=-0.19

400500600700800

Tota

l sum

mer

prec

ipita

tion Y=1572.7+0.49X

R=-0.104

1950 1960 1970 1980 1990 2000100200300400

Tot

al su

mm

er p

reci

pita

tion

Y=-972.3+0.6XR=0.16

Year

100200300400

June

Y=-733.7+0.45XR=0.12

0100200300

July

Y=-1393.3+0.77XR=0.25

400600800

1000

Aug

ust

Y=-2240.2+1.42XR=0.2

Fig. 5 Precipitation (mm) changes during 1951–2002 in the (a) upper, (b) middle and (c) lower Yangtze basin, in June, July, August.

the upper Yangtze basin is characterized by decreasing summer precipitation (with precipitation anomalies of <50 mm) (Fig. 4(b)). If one explores the changing trends of precipitation by dividing the whole Yangtze basin into three parts based on longitude (Table 1), one can see that different changing trends occurred in different seasons (Fig. 5). In the upper Yangtze basin, both the precipitation in June and August and the summer precipitation show a downward trend, and the precipitation in July has an upward trend. In middle Yangtze basin, precipitation in August shows a downward trend, while that in June, July and summer shows an upward trend. For the lower Yangtze basin, the precipitation in June, July, August and the total summer precipitation show an upward trend (Figs 3 and 4). Observed temperature trends Figure 6 demonstrates that, in June, the upper Yangtze basin is dominated by upward temperature trends at >90% significance level. Temperature in some places in the lower Yangtze basin is characterized by upward trends. The downward trends occurred

(c)

(b) (a)



73

Fig. 6 Spatial distribution of Mann-Kendall changing trends of temperature in (a) June, (b) July and (c) August at 90, 95 and 99% significance levels.

mainly in the northern middle Yangtze basin at >90% significance level. However, in July most places in the Yangtze basin are dominated by downward temperature trends, which are concentrated mainly in the middle Yangtze basin at >90% and even at >99% significance level. Temperature in August presents similar changing trends when compared to temperature changes in June; that is, the upward temperature trends dominate the upper Yangtze basin, and downward temperature trends occurred in the southeastern and northern Yangtze basin at >90% significance level. Figure 7 shows the temperature changes in the upper, middle and lower Yangtze basin, respectively. For the upper Yangtze basin, the temperature in June and August shows an upward trend, while the temperature in July shows a downward trend. For the middle and lower Yangtze basin, the temperature changes are dominated by downward trend: in the lower Yangtze basin only temperature in June shows an upward trend.

(a)

(b)

(c)

Qiang Zhang et al.


74

1950 1960 1970 1980 1990 2000

16

17

18

June

Y=-18.1+0.018XR=0.39

Yea

16

18

20

July

Y=22.3-0.002XR=-0.06

16

18

20

Aug

ust

Y=8.49+0.005XR=0.12

1950 1960 1970 1980 1990 200022232425

June

Y=36.44-0.007XR=-0.15 24

26

28

July

Y=45.2-0.01XR=-0.23

Year

24262830

Aug

ust Y=38.2-0.01X

R=-0.11

1950 1960 1970 1980 1990 20002425

26

27

June

Y=20.1+0.003XR=0.06

28

30

July

Y=42.4-0.01XR=-0.12

Year

26

28

30

32

Aug

ust

Y=51.14-0.01XR=-0.21

Fig. 7 Temperature (°C) changes during 1951–2002 in the (a) upper, (b) middle and (c) lower Yangtze basin in June, July, August.

Figure 8 demonstrates the summer temperature anomalies in the Yangtze basin. One can see from Fig. 8 that, during the summer, the middle Yangtze basin is domi-nated by decreasing temperature and the lower parts of the upper Yangtze basin are dominated by increasing temperature, which could be the result of increasing cloud coverage because of increased precipitation.

Fig. 8 Summer (June, July, August) precipitation anomalies between the 1990s and 1951–1990.

(a) (b)

(c)



75

PRECIPITATION, TEMPERATURE AND RUNOFF TRENDS IN THE UPPER, MIDDLE AND LOWER YANGTZE BASIN Mann-Kendall analysis is a non-parametric statistical test method for trend detection in a time series which does not assume the normality of distribution. Figure 8 shows the MK analysis of the monthly precipitation of the upper, middle and lower Yangtze basin. The UF curve shows the changing trend of a time series. The time series will be in downward trend if UF < 0 and vice versa. If the UF value is greater than the critical values (the two dashed lines beside the one showing zero value), then this upward or downward trend is at >95% significance level. When the UF and UB curves intersect at a certain time, the intersection point denotes the jumping time. The analysis shows that in the upper Yangtze basin (Fig. 9), during 1951–2000, the precipitation has a downward trend, which in 1980–1985 was significant at the 95% level. The jump in this time series occurred about 1995–1997. The changing trends of precipitation in the middle Yangtze basin are somewhat complex. An upward trend was found during 1980–1990. During the last 10 years, the precipitation in the middle Yangtze basin shows a downward trend. The whole time series of precipitation in the middle Yangtze basin has no obvious “jumps”. The precipitation in the lower Yangtze basin shows a downward trend during about 1951–1983. After 1983, the precipitation

-2

0

2

UF UB

UF

Year

Upper Yangtze Catchment

-2

0

2

UB

1950 1960 1970 1980 1990 2000

500600700800900

Prec

ipita

tion

(mm

) (u

pper

YC

)

-2

0

2

UF UB

UF

Middle Yangtze Catchment

-2

0

2

UB

1950 1960 1970 1980 1990 2000

400

500

600

700

800

Prec

ipita

tion

(mm

)

Year

-4

-2

0

2

4 UF UB

UF

-4

-2

0

2

4

UB

1950 1960 1970 1980 1990 2000

420

630

840

Prec

ipita

tion

(mm

)

Year

Lower Yangtze Catchment

Fig. 9 Results of Mann-Kendall analysis of monthly precipitation during 1951–2002 in the (a) upper, (b) middle and (c)lower Yangtze basin. The dashed lines, above and below the line denoting zero, denote critical values (significance level: 95%).

(b)(a)

(c)

Qiang Zhang et al.


76

-2

0

2

4 UF UB

UF

Upper Yangtze Catchment

-2

0

2

4

UB

1950 1960 1970 1980 1990 200016

17

18

Tem

pera

ture

( o C

)

Year

-2

0

2

4

UF UB

UF

-2

0

2

4

UB

1950 1960 1970 1980 1990 200023

24

25

Tem

pera

ture

( o C)

Year

Middle Yangtze Catchment

-2

0

2

UF UB

UF

Lower Yangtze Catchment

-2

0

2

UB

1950 1960 1970 1980 1990 2000

26

27

28

Tem

pera

ture

( o C

)

Year Fig. 10 Results of Mann-Kendall analysis of the monthly mean temperature during 1951–2002 in the (a) upper, (b) middle and (c) lower Yangtze basin. The dashed lines, above and below the line denoting zero, denote critical values (significance level: 95%).

shows an upward trend and during the last five years this trend is at >95% significance level. The jump time in the precipitation series in the lower Yangtze basin is during 1991–1993. Figure 10 shows the temperature trends in the upper, middle and lower Yangtze basin. It can be seen from Fig. 10 that the temperature in the upper Yangtze basin decreases during 1955–1994 and increases thereafter. The jump time of this time series lies in 1997. The temperature trend in the middle Yangtze basin is somewhat similar to that in the lower Yangtze basin. The temperature in the middle Yangtze basin shows a downward trend during 1951–2002. However, the temperature in the lower Yangtze basin, increases during 1959–1969 and decreases thereafter. Figure 11 demonstrates that the jump time of runoff changes at the Yichang station is around 1997. It can be seen clearly from Fig. 10 that during 1968–1998 runoff at the Yichang station is in downward trend. The changing trends of the runoff at Hankou and Datong stations are similar (location of these three stations can be found in Fig. 2). The beginning time of the upward trend of the runoff at Hankou and Datong stations is about 1980 and 1968 respectively (Fig. 11). This abrupt change in the runoff at Hankou and Datong stations may be the result of the abrupt change of the wind field in

(b)(a)

(c)



77

-2

0

2 UF UB

UF

Yichang station

-2

0

2

UB

1950 1960 1970 1980 1990 2000

30000

45000

Disc

harg

e (m

3 /s)

-2

0

2

4

UF UB

UF

-2

0

2

4

UB

Hankou station

1950 1960 1970 1980 1990 2000

30000

45000

60000

Dis

char

ge (m

3 /s)

-2

0

2

4 UF UB

UF

Datong station

-2

0

2

4

UB

1950 1960 1970 1980 1990 2000

38000

57000

76000

Dis

char

ge (m

3 /s)

Year (AD) Fig. 11 Results of Mann-Kendall analysis of the monthly mean discharge at (a) Yichang, (b) Hankou and (c) Datong stations (see Fig. 2 for locations). The dashed lines, above and below the line denoting zero, denote critical values (significance level: 95%).

the Asian monsoon region and above the west Pacific Ocean (Fu & Fletcher, 1988). The jump time of the runoff of these two stations is about 1985–1990, earlier than that of the Yichang station. FLOODS IN THE UPPER, MIDDLE, LOWER AND THE WHOLE YANGTZE BASIN, RESPECTIVELY Data of floods that occurred during 1950–2000 were collected. The disaster magnitude was determined by the combination of flood-affected area, amount of damage (popula-tion and houses, etc.), peak flood runoff and also the occurrence frequency of floods (the frequency of extremely large floods is below 5%, that of larger floods between 5 and 10% and that of usual floods 10–20%). Here all floods, including extremely large floods, larger floods and usual floods, were considered without regard for flooding magnitude. The number of floods is statistically calculated per 5 years (Fig. 12). Figure 12 demonstrates clearly that the floods in the upper Yangtze River are in downward trend. In the middle and lower Yangtze River floods are in upward trend. As for the whole Yangtze basin, the changing trend of the number of floods is complex. During 1950–1980, the floods are in downward trend, and afterwards in upward trend.

(b)

(c)

(a)

Qiang Zhang et al.


78

1950 1960 1970 1980 1990 20001234567

Year

Floods in the Yangtze Catchment

0

2

4

Num

ber o

f flo

ods

Floods in the middle and lower Yangtze River

0246 Floods in the upper Yangtze River

Fig. 12 Number of floods in the upper, middle and lower, and whole Yangtze basin, respectively. Floods presented in this figure were considered by the combination of flood-affected area, amount of damage (population and houses, etc.), peak flood runoff and also the occurrence frequency of floods.

It can be seen from Fig. 12 that, for the whole Yangtze basin and for the lower and middle Yangtze River, the floods are in upward trend, especially during 1985–2000. Global warming may exert a great impact on the probabilities of occurrence of floods in the Yangtze basin, which is in line with those of GCMs (Gao et al., 2001), but not confirmed in some areas, such as Central Europe (Mudelsee et al., 2003). The regional response to global warming varies within the Yangtze basin, especially for correlations between precipitation and temperature. SUMMARY AND CONCLUSIONS 1. Spatially interpolated precipitation and temperature trends in the Yangtze basin

indicate that an upward precipitation trend dominates the middle and lower Yangtze basin, and upward temperature trend occurs in the upper Yangtze basin. Analysis results of temperature and precipitation trends by dividing the whole Yangtze basin into three parts, namely upper, middle and lower, based on the longitude, indicated that summer (June, July, August) precipitation and precipita-tion in most individual months of the summer are in upward trend. However, in the upper and middle Yangtze basin, summer precipitation and precipitation in most months of summer are in downward trend. The main reason for this phenomenon may be that the middle and lower Yangtze River reaches are dominated by east Asian monsoon climatic changes and the upper Yangtze River is dominated by the Indian Monsoon climatic system. These two climatic systems are not in synchronization spatially and temporally, leading to different changing trends of precipitation and temperature in the upper, middle and lower Yangtze basin.

2. Mann-Kendall analysis of monthly precipitation in the upper, middle and lower Yangtze basin and of runoff (at Yichang, Hankou and Datong stations) indicates



79

that the trend of runoff is somewhat coincident with precipitation trend, but the changing characteristics are different. The upward trend of precipitation in the upper Yangtze basin begins at about 1980. The precipitation in the middle Yangtze basin is in downward trend after 1993. Upward precipitation trend in the middle Yangtze basin starts at about 1983. The runoff of Yichang, Hankou and Datong stations is in upward trend after 1998, 1980 and 1967, respectively. It should be mentioned that an upward trend of flood runoff over time is evident. Yichang station mainly collects the runoff from the upper Yangtze basin, while Hankou and Datong stations mainly collect the runoff from the upper, middle and lower Yangtze basin. Therefore, one can tentatively conclude from the runoff, the precipitation trends of the upper, middle and lower Yangtze basin and the spatially interpolated precipitation trends in the whole Yangtze basin that precipitation trend corresponds well to the runoff trends. Trend analysis shows that the middle and lower Yangtze basin are likely to face more serious flooding hazards.

3. Time series analysis of the occurrence of floods indicates that the floods in the upper Yangtze River are in downward trend and floods in the middle and lower Yangtze River are in upward trend. This result is further corroborated by evidence from analysis results of the monthly mean flood runoff and precipitation. As such, more attention should be given to the middle and lower Yangtze River in flood-controlling activities.

4. The temperature analysed in this paper is the Earth surface temperature, but not at high latitude (e.g. higher than the stratosphere). Therefore, cloud cover will influence the Earth’s surface, increased cloud coverage will decrease the ground surface temperature. This is the possible reason for the adverse spatially inter-polated trends between temperature and precipitation. It should be mentioned that only precipitation and temperature are considered as factors influencing runoff production. Evaporation, wind, solar radiation and human activities are not con-sidered in this paper. Human-induced deforestation and soil erosion will lead to higher water levels and intensify the flood hazards (e.g. the 1998 flood) (He & Jiao, 1998). Therefore, the mechanisms of the occurrence of floods in the Yangtze basin will be more complex over time because of intensified human activities (Yin & Li, 2001) and further research will be necessary.

Acknowledgements This paper is financially supported by the Key project of the Chinese Academy of Sciences (KZCX3-SW-331), the National Natural Science Foundation of China (Grant no. 40271112), the National Post-doctoral Foundation of China and the K.C. Wong Post-doctoral Education Foundation, Hong Kong. The authors extend their thanks to Prof. Tadeusz Niedzwiedz and an anonymous reviewer for their revision suggestions and advice. The authors would also like to thank the National Climate Centre in Beijing and CWRC of the Yangtze River in Wuhan for providing valuable climate and hydrological data sets. REFERENCES Boorman, D. B. & Sefton, C. E. M. (1997) Recognizing the uncertainty in the quantification of the effects of climate

change on hydrological response. Climate Change 35, 415–434. Buishand, T. A. (1982) Some methods for testing the homogeneity of rainfall records. J. Hydrol. 58, 11–27.

Qiang Zhang et al.


80

CWRC (Changjiang Water Resources Commission) (2000) Flood-controlling Handbook of the Yangtze River, 100–161. CWRC, Ministry of Water Resources, China (in Chinese).

CWRC (Changjiang Water Resources Commission) (2002) Flooding and Drought Hazards in the Yangtze River Basin. Chinese Water Resources and Water Power Press, Beijing (in Chinese).

Fu, C. B. & Fletcher, J. (1988) Larger signals of climatic variation over the ocean in the Asian monsoon region. Adv. Atmos. Sci. 5, 389–404.

Gao, X. J., Zhao, Z. C. & Ding, Y. H. (2001) Climate change due to greenhouse effects in China as simulated by a regional climate model. Adv. Atmos. Sci. 18(6), 1224–1230.

Gemmer, M., Becker, S. & Jiang, T. (2004) Observed monthly precipitation trends in China 1951–2002. Theor. Appl. Climatol. 77, 39–45.

Grossman, M. J. (2001) Large floods and climatic change during the Holocene on the Ara River, Central Japan. Geomorphol. 39, 21–37.

He, X. B. & Jiao, J. R. (1998) The 1998 flood and soil erosion in the Yangtze River. Water Policy 1, 653–658. Helsel, D. R. & Hirsch, R. M. (1992) Statistical Methods in Water Resources. Studies in Environmental Science, Elsevier,

Amsterdam, The Netherlands. Kendall, M. G. (1938) A new measure of rank correlation. Biometrika 30, 81–93. Libiseller, C. (2002) A program for the computation of multivariate and partial Mann-Kendall test. University of

Linköping, Sweden. Luo, C. Z. & Le, J. X. (eds) (1996) Chinese Floods, 400–433. Bookstore Press of China, Beijing (in Chinese). Mudelsee, M., Börngen, M., Tetzlaff, G. & Grünewahl, U. (2003) No upward trends in the occurrence of extreme floods in

central Europe. Nature 425, 166–169. Muzik, I. (2002) A first-order analysis of the climatic change effect on flood frequencies in a subalpine watershed by

means of a hydrological rainfall–runoff model. J. Hydrol. 267, 65–73. Prudhomme, C., Jakob, D. & Svensson, C. (2003) Uncertainty and climatic change impacts on the flood regime of small

UK catchments. J. Hydrol. 277, 1–23. Simonovic, S. P. & Carson, R. W. (2003) Flooding in the Red River basin—lessons from post flood activities. Natural

Hazards 28, 345–365. Wei, F. Y. (1999) Diagnostic Technology of Modern Climatic Statistics, 69–71. Meteorological Press, Beijing (in

Chinese). Yin, H. F. & Li, C. A. (2001) Human impact on floods and flood disasters on the Yangtze River. Geomorphol. 41, 105–

109. Zhou, Y. K., Ma, Z. Y. & Wang, L. C. (2002) Chaotic dynamics of the flood series in the Huaihe River Basin for the last

500 years. J. Hydrol. 258, 100–110. Received 29 March 2004; accepted 2 November 2004

Documents

Precipitation, temperature and runoff analysis from 1950 ...precipitation leads to increasing cloud coverage and, hence, results in decreasing ground surface temperature. Average monthly