As the global atmospheric circulation abruptly changed on a planetary scale in the late 1970s, ^{[1, 2]} regional climate changed a lot, which resulted in serious floods in south China and severe droughts in the north.^{[3, 4]}

Northeast China is regarded as one of the largest commodity grain bases and most potential agriculture production areas in China.^[5] It is also believed that summer is the primary season for northeast crops to grow, with frequent precipitation which affects the yield significantly as well as its distribution. Thus, the research into the summer precipitation is fairly crucial.

Lian et al.^{[6, 7]} set up the standard of generation and retreat of East Asian summer monsoon (EASM) and ascertained the corresponding index in northeast China, after studying the influence of EASM on summer precipitation in the northeast. They found that before and after EASM generated in (retreated from) the northeast, the precipitation increased (decreased) substantially. Meanwhile, the intensity index was positively correlated with the precipitation in July and August over most parts of the northeast. Shen et al.^[8] used the observational data from 79 stations in the northeast and the 40-yr ECMWF Re-Analysis (ERA-40) data, and studied the characteristics of large-scale circulation affecting the interannual variation of precipitation in all summer months over this region. Results showed that a cold vortex was prominent in affecting the precipitation in June, EASM in July, while the summer monsoon and mid-high-latitude northeast Asian were blocking together in August. Cui^[9] chose the data between 1981 and 1986 from 18 air sounding stations to study the moisture input and output of all edges of the northeast, and to analyze the water budget above this region. Results showed that the moisture input was larger than the output around the whole year in northeast China, and, summer was the only season in which the input was much greater than the output. However, discrepancy may partly be spurious because the number of stations was small, thereby leading to the spatial inaccuracy of such data unable to meet the requirements, and a large number of data processed in a simple way differing from the reality. Gao^[10] analyzed National Centers for Environmental and Prediction (NCEP) daily reanalysis data from 1948 to 2006 to review the precipitable water in the northeast China, the spatial characteristics of meridional and zonal transport of moisture, total moisture transport, and moisture divergence flux, and interannual and interdecadal variation of surrounding moisture transport. He ultimately discovered that the precipitation appeared on a diminishing scale from the south to the north of northeast China. As we can see from the above studies, meteorologists were aware of the importance of the water cycle early enough; however, a great many studies were still only on single element characteristics of the water cycle, which were lacking in comprehensive study of the elements; besides, some work was accomplished a long time ago, excluding recent data or any study on interrelation of these elements. Lately, ECMWF has released abundant meteorological data whose accuracy is much improved, ^{[11, 12]} enabling us to compute and study the characteristics of all elements and the water budget in northeast China in a comprehensive, systematic, and quantificational way. Based on this, this paper discusses the relation of all elements of the water cycle and the water budget relationship in northeast China, attempting to provide more accurate reference information for summer precipitation predictions over the northeast based on our existing summer rainfall prediction methods.^{[13– 23]}

This paper employs the ERA-Interim data, where the elements include u and v wind fields and specific humidity field that were measured four times a day and eight vertical layers, which are 1000 hPa, 925 hPa, 850 hPa, 700 hPa, 600 hPa, 500 hPa, 400 hPa, 300 hPa, respectively; monthly average precipitation and evaporation from 1979 to 2010, with horizontal resolution of 2.5° × 2.5° . In this paper, the precipitable water and horizontal moisture flux divergence respectively represent the mean values of their corresponding values of spring (March– May), summer (June– August), autumn (September– November), and winter (December– next February). The precipitation and evaporation accounts for their total amounts of each season (three months per season). In addition, the gridpoint precipitation data from ECMWF reflects the observational precipitation in different seasons from the stations in the northeast more clearly. We choose such data for further computation and procession. Full-layer precipitable water, full-layer zonal and meridional moisture flux are the integrals of 300 hPa from above the terrain.^[24] The equation to compute full-layer atmospheric precipitable water is

the equations to compute full-layer zonal and meridional moisture flux are respectively

the equation to compute horizontal moisture flux divergence is

where

Here, Q is the atmospheric precipitable water of the full layer integration, Q_u and Q_v are respectively the zonal and meridional moisture flux of the full layer integration, MC is the horizontal moisture flux divergence, and Q is the horizontal moisture flux vector of the full layer integral, g is the gravitational acceleration (taking 9.8 kg· m· s^{− 2}), p_s is the surface atmospheric pressure, q is specific humidity, u and v are the zonal and meridional wind speed respectively.

Moisture flux divergence represents the moisture converging to or diverged from per unit volume in per unit time. The fact that moisture input exceeds the output (divergence is negative) is referred to as “ moisture convergence” , indicating that the convergence of moisture advected from the periphery towards this area, and such an area is called a “ moisture sink” ; the fact that moisture input is less than the output (divergence is positive) is referred to as “ moisture divergence” , indicating that the divergence of moisture advected from this area towards the periphery, and such an area is called a “ moisture source” . As a result, the spatial distribution of moisture flux convergence (divergence) can be used to describe that of the “ moisture sink (moisture source)” . Moisture flux divergence in this paper is calculated by integrating the full-layer moisture flux above this terrain and then computing divergence.

From Fig. 1, precipitable water (Fig. 1(a)), precipitation ((Fig. 1(e)), and evaporation (Fig. 1(i)) in spring in the southeast are more than in the northwest. The maximum precipitation and evaporation occur in the area adjacent to the Liaoning Peninsula where moisture flux divergence (Fig. 1(m)) changes into convergence. The precipitable water (Fig. 1(b)), precipitation ((Fig. 1(f)), and evaporation (Fig. 1(j)) in summer are greater than in spring, with the area under the precipitation and evaporation of high values extending, and the moisture convergence intensified around areas near the Liaoning Peninsula (Fig. 1(n)). Compared the case in summer, the precipitable water (Figs. 1(c) and 1(d)), precipitation ((Figs. 1(g) and 1(h)), and evaporation (Figs. 1(k) and 1(l)) in autumn and winter are less, characterized by the fact that such values in the southeast are greater than in the northwest in general. For moisture flux divergence, it plays a dominant role in the northeast of China in autumn (Fig. 1(o)), while neither convergence nor divergence occurs conspicuously in winter (Fig. 1(p)). Overall, a strong convergence zone is identical with the area of maximum precipitation and evaporation in summer in the northeast of China.

Fig. 1.

Figure Option ViewDownloadNew Window

Fig. 1. The spatial distributions of water cycle elements in northeast China in different seasons. Panels (a)– (d) represent the spatial distributions of the atmospheric precipitable water in spring, summer, autumn, and winter in mm; panels (e)– (h) are similar to panels (a)– (d), only except that they represent the precipitation in mm; panels (i)– (l) are similar to panels (a)– (d), only except that they represent the evaporation in mm; panels (m)– (p) are similar to panels (a)– (d), only except that they represent the moisture flux divergences in 10− 5 mm.

Figure 2 shows the interannual variations of water cycle elements in northeast China in different seasons, in which the precipitable water, precipitation, and evaporation are the mean values in this area. From Fig. 2 and Table 1, the precipitable water (Fig. 2(a)) and precipitation (Fig. 2(e)) do not change drastically, the evaporation (Fig. 2(i)) increases and moisture flux divergence (Figs. 2(f) and 2(g)) goes downwards, each with no significant variation. In summer and autumn, the precipitation (Figs. 2(f) and 3(g)) in the northeast of China trends downward distinctly, which is in accordance with the results of Huang Gang^[25] and Liu Yong, ^[26] moisture flux divergences (Figs. 2(n) and 2(o)) are on a rise. Precipitable water (Figs. 2(b) and 2(c)) and evaporation (Figs. 2(j) and 2(k)) do not seem to be obvious. In winter, the evaporation (Fig. 2(l)) decreases, the variations of precipitable water (Fig. 2(d)), precipitation (Fig. 2(h)), and moisture flux divergence (Fig. 2(p)) are all fairly small. In general, the variations of moisture flux divergence in the northeast all through the year resembles that of precipitation. When precipitation declines, moisture flux divergence goes upward, leading to the moisture convergence weakening and divergence intensifying, which together contribute to the disadvantage of precipitation.

Fig. 2.

Figure Option ViewDownloadNew Window

Fig. 2. The interannual variations of water cycle elements in northeast China in different seasons (1979– 2010). Panels (a)– (d) represent the interannual variations of the precipitable water in spring, summer, autumn, and winter in mm; panels (e)– (h) are similar to panels (a)– (d), only except that they represent precipitation in mm; (i)– (l) are similar to (a)– (d), only except that they represent evaporation in mm; panels (m)– (p) are similar to panels (a)– (d), only except they represent moisture flux divergence in 10− 5 mm. The curves represent the interannual variation lines, the dotted lines denote the trend lines; the straight lines refer to the mean values.

Table 1.

Table 1.

Table 1. Trend coefficients of water cycle factors in different seasons. Atmospheric precipitable water Precipitation Evaporation Moisture flux divergence Spring 0.10 0.10 0.26 − 0.21 Summer − 0.15 − 0.36a 0.13 0.21 Autumn 0.02 0.37a − 0.03 0.23 Winter − 0.15 0.01 − 0.29 − 0.04 a represents a 95%-confidence level.

Table 1. Trend coefficients of water cycle factors in different seasons.

In this paper the ERA-Interim data from ECMWF is used to analyze the spatial distribution and temporal variation of the precipitable water, precipitation, evaporation, and divergence of moisture flux of northeast China and the water balance of that area. From the present study we draw some conclusions as follows.

(i) The spatial distributions of precipitable water, precipitation, and evaporation feature that the values of these elements in the southeastern area are all larger than those in the northwestern area; in summer, much precipitation and evaporation occur in the Changbai Mountain region as a strong moisture convergence region; in spring and autumn, moisture divergence dominates the northeast China; in winter, the moisture divergence and convergence are weak in this area.

(ii) From 1979 to 2010, the total amount of precipitation of summer and autumn in northeast China decreased markedly; especially from 1999 to 2010, the summer precipitation was always shown as a negative anomaly. Additionally, other elements in different seasons change in a truly imperceptible way.

(iii) In spring, evaporation exceeds precipitation in northeast China with moisture output; in summer, the precipitation is more prominent with moisture input; in autumn and winter, the precipitation plays a more dominant role in the northern part of northeast China, featuring that moisture input exists, but the evaporation exceeds the precipitation in the southern part with moisture output occuring; the closer to the sea the areas, the greater the difference between evaporation and precipitation will be.

(iv) ERA-Interim data has properly described the water balance between different seasons in northeast China. Based on ERA-Interim data, the moisture sinks computed through moisture convergence and moisture local variation are quite consistent with those computed through precipitation and evaporation, which proves that ERA-Interim data can be used in studying the water balance in northeast China.

(v) On a seasonal scale, the moisture convergence has a greater influence than the local moisture variation on the moisture sink, and the latter is variable slightly, generally it is a constant. Likewise, in different seasons, the total precipitation has a much greater influence on moisture sink than the evaporation.

Although ERA-Interim data can properly describe the moisture balance of different seasons in northeast China, there exist some deviations. This could be caused by the embodiment of the assimilation system for evaporation. Because of the complexity of the land surface process, the analysis data do not include enough information about the land-atmospheric exchange, which leads to a larger uncertainty of the estimation of evaporation.^{[28, 29]} This problem should be noticed in the research of regional water cycle balance.