CHANGES OF PAN EVAPORATION IN THE UPPER REACH OF THE YANGTZE RIVER*

RONG Yan-shu

College of Hydrology and Water Resources, Hohai University, Nanjing 210098, China,

E-mail: ysron@hhu.edu.cn

WANG Wen

State Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering, Hohai University, Nanjing 210098, China

JIANG Hai-yan

College of Hydrology and Water Resources, Hohai University, Nanjing 210098, China

CHANGES OF PAN EVAPORATION IN THE UPPER REACH OF THE YANGTZE RIVER*

RONG Yan-shu

College of Hydrology and Water Resources, Hohai University, Nanjing 210098, China,

E-mail: ysron@hhu.edu.cn

WANG Wen

State Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering, Hohai University, Nanjing 210098, China

JIANG Hai-yan

College of Hydrology and Water Resources, Hohai University, Nanjing 210098, China

The upper reach of the Yangtze River, 4 511 km long from west to east, contains a great amount of water resources of the Yangtze River Basin. This article studies the characteristics of the pan evaporation, the related meteorological variables, and their effects on the pan evaporation, based on the data of the daily pan evaporation (1980-2008) and other meteorological variables (1961-2008). The results show that the linear trend of the pan evaporation has remarkable regional features, i.e., the decrease trend in the southwest and the increase trend in the northeast of the investigated region, and the Yangtze River is approximately the boundary of these trends. The meteorological variables have different effects on the pan evaporation depending on the fact that they are in the category the thermal variables or the dynamic variables. The thermal meteorological variables (i.e., air temperature, diurnal temperature range, and sunshine duration) have positive partial correlations with the pan evaporation, while the dynamic ones (air pressure, rainfall, and relateive humidity) have negative correlations with the pan evaporation. The correlation of the wind speed remains to be investigated.

linear time trend, meteorological variable, pan evaporation, partial correlation, Yangtze River

Introduction

The fact of the global warming has been accepted for about twenty years[1-3]since it was observed that the global mean surface air temperature has increased 0.74oC during the recent hundred years according to IPCC AR4[4]. One of the consequences of global warming is that the air near the surface becomes drier, which would be accompanied with an increase of the evaporation from terrestrial open water bodies. However, despite the observed increase in average temperature, the observations in many regions of the world should that the pan evaporation at a regionalscale has been steadily decreasing over the past 50 years[5-8].

In India, both the pan evaporation and the potential evapotranspiration in different seasons had decreased according to the different time scale data[6]. Roderick and Farquhar[8]found that the pan evaporation in New Zealand decreased on average by 4.3 mm/a2during the period from 1970 to 2002. The similar decrease of the pan evaporation was also observed in Canada[9], in the Tibetan plateau[10], in New Zealand[8], in the Huaihe River Basin in China[11], in Japan[12], and in British Isles[13].

Most of the researches in the past focus on the survey of some regions and the rate of the pan evaporation reductions and on their relations with climate changes and the global hydrological cycle. In some studies the reduction was explained by concurrent decreases in the global radiation[7,13]. But in other studies the changes were attributed to the deficit of vapor pressures and the

decrease of windspeeds[14,15]. Peterson et al.[5]pointed out that the downward trend in the pan evaporation over most of the United States and the former Soviet Union has a good correspondence with the decreases both in the maximum summer temperature over these regions and in growing-season degree-days. On the basis of regression analysis, Chattopadhyay et al.[6]indicated that the increasing relative humidity is strongly related to the overall decrease in the pan evaporation. Roderick and Farquhar[7], Cohen and Stanhill[14]explained that the decrease of the pan evaporation is due to the increasing cloud coverage and aerosol concentration.

Table 1 The number of stations with the trends of decrease and increase ofEpan

In China, similar decreases of the pan evaporation were reported in the past few decades, such as in the Yangtze River Basin, the Yellow River Basin, and the Huaihe River Basin. The major reasons were considered as the decrease of solar radiation, sunshine duration, wind speed, diurnal temperature range[11,16-19].

The upper reach of the Yangtze River is selected as the investigated region in the present study, and this basin originates from the central of the Qinghai Tibetain Plateau, and expands 4 511 km towards east up to the central China (24oN-36oN, 90oE-112oE), including eight rivers, seven provinces and one city. The historic changes of the pan evaporation and the linear time trends over the past few decades are presented based on 66 stations located in different regions of the upper reach of the Yangtze River. By means of a partial correlation analysis, the most dominant meteorological variables related with such changes are identified and discussed.

1. Data

All the data include two aspects: the historic meteorological variables and the pan evaporation. The meteorological variables, during about fifty years (1961-2008), are used in the present study to work out their trends in year and seasons, and the variables include the mean air pressure (PA), the mean air temperature (TA), the diurnal temperature range (TR), the monthly rainfall (R), the mean monthly relative humidity (RH), the monthly sunshine duration (SD), the mean monthly wind speed (WS), as obtained from the 90 well-distributed stations shown in Fig.1. The duration of the seasons is defined as winter (December, next January and February, DJF), spring (March to May, MAM), summer (June to August, JJA), and autumn (September to November, SON).Φ20 evaporation pan, of 0.20 m in diameter, 0.10 m deep, and placed on the ground. The pan evaporation anomalies are calculated on the basis of the normal climate period data.

Fig.1 Location of meteorological stations in the upper reach of the Yangze River

The 66 stations for the observations of the pan evaporation, selected from the 90 stations shown in Fig.1, cover various regions in the upper reach of the Yangtze River. All meteorological variables mentioned above are standardized.

With respect to the pan evaporation (Epan), the data of thirty years (1980-2008) are recorded from

2. Results and discussions

2.1Linear time trends of Epan

Table 1 is the number of the stations with the trends of decrease and increase of the pan evaporation in year and seasons in the investigated regions, meanwhile, the stations with statistical significance (p) of less than 0.05 is listed in this table. It should be noted that among the 66 stations, there are about or larger than 70% of stations, which produce data for lineartime trends of Epanin the durations of spring, summer and autumn, with p＜0.05. In winter, only 56% of stations produce data for the trend with statistical significance, and for the annual trend, the ratio is only 13.64%. The results show that for those 66 stations the linear time trends of Epanin four seasons are significant but not for the annual trend.

For the linear time trend of Epan, there are significant differences for seasons for the 66 stations. In spring and summer, about 60% of stations give the trend of the decrease of Epan, while it is the trend of the increase of Epanfor about 86% and 66% of stations in autumn and winter, respectively. The ratio of decrease to increase is 29/37 for the annual trend of Epan.

Figure 2 displays the distribution of linear time trends of Epanfor year and four seasons, in which the numbers express the change of the pan evaporation (mm/10a), the positive values for the increasing trend, the negative ones for the decreasing trend, and the shaded areas mean statistical significance of p < 0.05. It can be seen that the strongly regional features of Epancan be seen in data for year and seasons. The zero isoline of the linear trend divides the investigated region into two parts. The trend of the decrease of Epanappears in southwest and the trend of the increase in northeast and the Yangtze River is roughly their boundaries.

For the annual trend in Fig.2(a), the increase of Epanoccurs in Sichuan, Qinghai, Gansu, and the headstream of the Yangtze River, the maximum increase is about 17 mm/10a, while the largest decrease of Epanreaches –32 mm/10a, in the region near Yunnan province. It could be observed that the decrease trend of Epanis more significant than the increase one.

Fig.2 Distributions of linear time trend of Epan(Dashed line for decrease trend, solid line for increase trend, shaded for p＜0.05 level)

In spring (Fig.2(b)) and summer (Fig.2(c)), the locations of the stations and the magnitudes of the trends of the decrease of Epanare highly similar, being 41 and 39 (62.12% and 59.09% in Table 1), respectively. Meanwhile, it could be noticed that the decreasing values of Epanreach –179 mm/10a and –97 mm/10a in those two seasons. Surely, there are parts of stations (37.88% and 40.91% for spring and summer) with the increasing trends of Epan, and theirmaximum values are only 37.4 mm/10a and 49.4 mm/10a. Comparisons of the absolute magnitudes of the changes between decrease and increase trends ofEpanindicate that the changes of the pan evaporation are mainly characterized by the decreasing trend in these two seasons.

In winter (Fig.2(e)), the zero isoline of the linear trend moves towards southwest, and then the linear trend could be characterized by the increase ofEpanin most of the stations (66.67%). It can be seen from this figure that there are still 22 stations (33.33%) with the decrease ofEpan, and the maximum decreasing value reaches –77.4 mm/10a and is larger than the absolute value of the increase rate of 44.4 mm/10a.

In autumn (Fig.2(d)), there is a remarkable difference in the number of the stations with significant changes ofEpanas compared with the other three seasons. Most of the stations (86.36%) show the increasing trend ofEpan, although the regional characteristics are of the decreasing trend in southwest and the increasing trend in northeast. Meanwhile, it could be noted that the values of the increase and decrease ofEpanare 61.0 mm/10a and –32.9 mm/10a, respectively.

2.2Effects of meteorological variables on Epan

There are many meteorological factors to be considered with respect to the effects on the pan evaporation. The key factors are the mean air pressure (PA), the mean air temperature (TA), the diurnal temperature range (TR), the monthly rainfall (R), the mean monthly relative humidity (RH), the monthly sunshine duration (SD), and the wind speed (WS). These factors could be divided into two categories: the thermal factors, includingTA,TRandSP, the dynamic ones,PA,R,RH, andWS. Figure 3 demonstrates the distributions of partial correlations between the pan evaporation (Epan) and these meteorological variables (PA,TA,TR,R,RH,DSandWD).

In order to investigate the effects of these meteorological factors onEpan, an analysis of partial correlations is made, to measure the degree of association between two random variables. The results of partial correlations between various meteorological factors andEpan, on the basis of the data of 66 stations, are shown in Fig.3, in which the gray areas stand for the negative partial correlations between meteorological variables andEpan, and white areas the positive partial correlations, while the numbers are their coefficients of the correlations.

Fig.3 Partial correlation between pan evaporation and meteorological elements (shaded for negative partial correlation)

It could be clearly observed from the gray area and the value of the correlation coefficient in Fig.3 that these seven meteorological variables in the correlations have different effects on the pan evaporation. Some variables, such asTA(Fig.3(b)),TR(Fig.3(c)), andSD(Fig.3(g)), have mainly positive partial correlations, while the other ones, such asPA(Fig.3(a)),RH(Fig.3(d)), andR(Fig.3(e)), have negative partial correlations. The correlation characteristics of the variableWS(Fig.3(f)) can not be determined directly due to the smallness of the coefficient of correlation. As a whole, those variables with the positive or negative partial correlations are approximately the thermal or dynamic variables, respectively, as well as the variableWS, according to the classification of the meteorological factors mentioned above.

In order to identify the dominant meteorological variables associated with the change ofEpan, the linear trends of bothEpanand the seven meteorological variables are compared for 66 stations (1980-2008). The ten stations with the maximum increasing or decreasing trends are listed in Table 2 and Table 3. The shaded boxes in the two tables indicate the negative partial correlations between those variables andEpan, and others the positive partial correlations.

In Table 2, for example, let us consider the effect of variablePAon the increase ofEpan. ThePAof the seven stations has negative correlations withEpan, that is to say,Epanincreases whenPAdecreases, at the same time, the value ofPAat the seven stations are all negative, i.e.,PAcontributes to the decrease ofEpan. Meanwhile, for those ten stations, there are completely negative correlations betweenRHandEpanand betweenRandEpan, which take all negative values. VariablesTA,TRandSDhave positive correlations withEpan, and take basically positive values.Epanincreases as those variables increase. It could be noted thatWSappears to have a positive correlation with the maximum increasing trend ofEpan.

For the ten stations in Table 3, the variables have similar features as those in Table 2.PAhas the negative correlation withEpan, with most of their values positive. As a result,Epandecreases whenPAincreases in the stations with the negative correlation. The variablesRHandRhave also negative correlations withEpanfor 70% of the stations, with large positive values leading to a sharp decrease ofEpan.TA,TRandSDhave positive correlations withEpan, that is, the less those variables, the large the decrease ofEpanwill be.

3. Conclusions

The linear trends ofEpanare investigated and the effects of the seven meteorological variables onEpanare analyzed for the upper reach of the Yangtze River, on the basis of the meteorological data (1961-2008) and the pan evaporation data (1980-2008). There are remarkable regional features in the trends ofEpan, that is, the decrease trend of the pan evaporation occurs mainly in the southwest, the increase trend in the northeast. The Yangtze River is approximately the boundary of the decrease/increase trends based on the results from either annual or seasonal data except for autumn.

The meteorological variables, includingPA,TA,TR,R,RH,SDandWS, have notable effects onEpan. Those variables may be divided into two catogories: the thermal variables and the dynamic variables, and they have different effects onEpan. The partial correlation analysis shows that the thermal variables (i.e.,TA,TRandSD) have positive correlations withEpan, that is to say,Epanincreasesas the trends of the thermal meteorological variables increase, while the dynamic variables (PA,R,RH) have negative correlations withEpanexcept for the wind speed (WS). The effect ofWSonEpanshould be investigated further.

Table 2 Ten maximum increasing linear trends inEpanand its corresponding variables

Table 3 Ten maximum decreasing linear trends inEpanand its corresponding variables

[1] THOMAS R. K., KEVIN E. T.Modern global climate change[J].Science,2003, 302(5): 1719-1723.

[2] JOHN H.Global warming–The complete briefing[M]. Cambridge, UK: Cambridge University Press, 1994.

[3] RONG Yan-shu. The characteristics of evolution of potential evaporation in Tianjin region of China[J].Journal of Hydrodynamics, Ser. B,2004, 16(3): 290-294.

[4] Intergovernmental Panel on Climate Change.Climate change 2007: The physical science basis (Summaryfor policymakers)[M]. Cambridge, UK: Cambridge University Press, 2007.

[5] PETERSON T. C., GOLUBEV V. S. and GROISMAN P. Y. Evaporation losing its strength[J].Nature,1995, 377(26): 687-688.

[6] CHATTOPADHYAY N., HULME M. Evaporation and potential evapotransporation in India under condition of recent and future climate change[J].Agriculture and Forest Meteorology,1997, 87(1): 55-73.

[7] RODERICK M. L., FARQUHAR G. D. The cause of decreased pan evaporation over the past 50 years[J].Science,2002, 298(15): 1410-1411.

[8] RODERICK M. L., FARQUHAR G. D. Changes in New Zealand pan evaporation since the 1970s[J].International Journal of Climatology,2005, 25(19): 2031-2039.

[9] BURN D. H., HESCH N. M. Trends in evaporation for the Canadian Prairies[J].Journal of Hydrology,2007, 336(1-2): 61-73.

[10] XU J., HAGINOYA S. and SAITO K. et al. Surface heat balance and pan evaporation trends in Eastern Asiain the period 1971-2000[J].Hydrological Process,2005, 19(11): 2161-2186.

[11] RONG Yan-shu, ZHOU Yun and WANG Wen. Changes in the Huaihe River Basin pan evaporation[J].Advances in Water Science,2011, 22(1): 15-22(in Chinese).

[12] JUN A. Long-term trend of pan evaporation measurements in Japan and its relevance to the variability of the hydrological cycle[C].Symposium on Water Resource and its Variabilityin Asia.Osaka, Japan, 2004, 37- 40.

[13] GERALD S., MARKUS M. Evaporative climate change in the British Isles[J].International Journal of Climatology,2008, 28(9): 1127-1137.

[14] COHEN S., STANHILL G. Evaporative climate changes at Bet-Dagan, Israel, 1964-1998[J].Agricultural and Forest Meteorology,2002, 111(2): 83-91.

[15] DONALD H. B., NICOLE M. H. Trends in evaporation for the Canadian Prairies[J].Journal of Hydrology,2007, 336(1-2): 61-73.

[16] WANG Yan-jun, JIANG Tong and XU Cong-yu. Spatial temporal change of 20 cm pan evaporation over the Yangtze River Basin[J].Advances in Water Science,2006, 17(6): 830-833(in Chinese).

[17] GUO Jun, REN Guo-yu. Recent change of pan evaporation and possible climate factors over the Huang Huai Hai watershed, China[J].Advances in Water Science,2005, 16(5): 666-672(in Chinese).

[18] ZOU Hong-chao, BAO Yan and ZHANG Cun-jie et al. An analytic and numerical study on the physical meaning of pan evaporation and its trend in recent 40 years[J].Chinese Journal of Geophysics,2006, 49(3): 680-688(in Chinese).

[19] RONG Yan-shu, TU Qi-pu and DUAN Li-yao. Characteristics of the interannual and interdecadal timescales of evaporation in northern China in the recent 115 years[J].Journal of Hydrodynamics, Sre. B,2006, 18(5): 593-596.

April 3, 2011, Revised May 30, 2011)

* Project supported by the National Natural Science Foundation of China (Grant Nos. 40771039, 50879017), the Ministry of Science and Technology (Grand No. 2008BAB29B08-02).

Biography: RONG Yan-shu (1961-), Female, Ph. D., Professor

2011,23(4):503-509

10.1016/S1001-6058(10)60142-4