Results and Discussion

Observed annual discharge changes at the six gauges during the study period are shown in Table 1. There are significant decreasing trends for all the gauges except for Tangnaihai gauge. The annual discharge at the Toudaoguai gauge is less than that at Lanzhou gauge and the discharge at Lijin gauge is less than that at Huayuankou gauge, suggesting the runoff absorbing is more than runoff generating in the Lanzhou-Toudaoguai and Huayuankou-Lijin sections.

Table 2 lists the climatic change in the Yellow River basin from 1960 to 2000 and the vegetation condition change from 1981 to 2000. The precipitation shows a decreasing trend with a significance level of 6%. Taking a significance level of 0.1, there is no significant trend in reference evapotranspiration. The decreasing in the river discharge possibly responds to the decrease in precipitation. There are no significant trends in relative humidity. Decreasing sunshine duration and cloud amount trends are found. The cloud amount decrease trend was also reported over much of China by Kaiser (1998). The Yellow River did become warmer according to the increasing trends in temperatures. The averaged LAI value over the Yellow River basin shows no significant trend, but the LAI value in irrigation districts (IDs) increases with a significance level of 13%. This indicates vegetation condition in the IDs became better from 1981 to 2000. The water consumptions and irrigation area increase from 1980 to 1995 confirms the vegetation condition change (Liu and Zhang, 2002).

Figure 2(a-e) shows the spatial distribution of relative trend magnitudes in precipitation, reference evapotranspiration, relative humidity, sunshine duration, and cloud amount over the Yellow River basin with a significant level of 5%. Figure 2(f) shows relative trend magnitude in LAI value over the basin and the lower reach IDs, with the marked zone of 20% significant level. The spatial pattern of precipitation trend shows increase in part region before the Tangnaihai gauge and decrease over the Loess Plateau. It can explain why there is no discharge decreasing trend at Tangnaihai gauge but decreasing trends at the other gauges. Reference evapotranspiration increases in the eastern Tibetan Plateau and Qingtongxia district, suggesting the evaporative demand of the atmosphere becomes larger in these regions. The relative humidity decreasing trend is obvious in the Qingtongxia-Hetao district, showing that region might become drier. Reduction in sunshine duration is observed in the lower reaches. There is decreasing cloud amount trend over the basin as mentioned before. LAI values show decreasing trends on the Tibetan Plateau and headwaters of tributaries at middle reaches. The vegetation degradation in the source region was also reported by Feng et al. (2005). And they indicated the degradation was enhanced by human activities. The LAI values in Qingtongxia-Hetao and lower reach IDs show obvious increasing trends, which are consistent with the irrigation area increase. The large discharge decreasing trends at the Toudaoguai and Lijin gauges should respond to the vegetation improvement and consequential water consumption in the Qingtongxia-Hetao and lower reach IDs.

The trend magnitudes of temperatures and DTR are shown in Figure 3 with a significant level of 5%. There are grand increasing trends in temperatures over the whole basin, except for the Sanmenxia-Huayuankou section. The mean temperature increase magnitude is large in the Tibetan Plateau and Qingtongxia-Hetao district. Minimum temperature shows increasing trends in the Tibetan Plateau and Qingtongxia-Hetao district, where cloud amount shows decreasing trends. Increases in cloud amount have been offered as a possible explanation for increasing minimum temperatures in other parts of the world (Kaiser, 1998). However, it seems other mechanisms should be considered in the Yellow River. Maximum temperature shows increasing trend over most of the basin except for one part of the lower reach district. In the Tibetan Plateau and Qingtongxia-Hetao district, the maximum temperature increasing magnitudes are obviously less than that of minimum temperature. A narrowing of the DTR is found in the Tibetan Plateau, Qingtongxia-Hetao district and lower reach district. The narrowing DTR in the Tibetan Plateau and Qingtongxia-Hetao district is due to differential changes in daily maximum and minimum temperatures. This is consistent with the global DTR trend (Easterling et al., 1997). The widening DTR is observed in the northern Loess Plateau, where has increasing trend in maximum temperature but not in minimum temperature.