*4.1. Analysis of Hydrometeorological Elements* 4.1.1. Trend Analysis

Overall, the trends of annual *Q*, *ET*, *P* and runoff coefficient (a) detected by the M–K trend test are summarized in Table 3. The study period is 50 years, and the significance level of 0.05 is ±1.96. If the M–K method statistic of *Q* is greater than 1.96 or less than −1.96, it indicates that the increase or decrease in *Q* is significant. Otherwise, the change is not significant, and the sign of the M–K method statistic represents the increase or decrease.


**Table 3.** M–K test for hydrometeorological elements in the Weihe River basin.

*Q* change showed a significant decreasing trend, with an M–K test statistic of −2.05. *ET* showed a non-significant increasing trend, with an M–K method statistic of −0.04. *P* showed a weak increasing trend, with an M–K method statistic of 0.61. Runoff coefficient showed a significant decreasing trend and the M–K method statistic is −3.29, as detailed in Table 3. From about 1990 to the early 2000s, *Q* decreased significantly compared with the previous period (Figure 2). The decrease in *P* at this stage is a factor, but the decline rate of *P* is much lower than that of *Q*. Therefore, the influence of underlying surface change on runoff may increase sharply, and the underlying surface becomes the most important factor to runoff. The underlying surface factors include terrain, soil, etc., among which human activities and vegetation are most important.

**Figure 2.** Trend of hydrological elements in the Weihe River basin. (**a**) Runoff Depth. (**b**) Potential Evapotranspiration. (**c**) Precipitation. (**d**) Runoff Coefficient.

#### 4.1.2. Mutation Point Detection

The M–K method was applied to detect mutation points for runoff depths from 1970 to 2019 in the Weihe River basin in Figure 3. Ufk is obtained from the M–K trend test. If Ufk is greater than the significance level, it indicates that the change is significant. Ubk is obtained by arranging the studied sequences in reverse order and using the M–K trend test. The significance level of 0.05 is ±1.96. If Ufk and Ubk intersect and are at the significance level, the intersection is likely to be a mutation point.

**Figure 3.** Detection of M−K mutation points of annual runoff depth in the Weihe River basin.

There are three intersections of Ufk and Ubk curves, the first two intersections are in the pre-series period and therefore excluded, and the third point is around 1990 and at the 0.05 significance level. Therefore, 1990 is likely to be the onset of the mutation.

#### *4.2. Analysis of Hydrometeorological Elements*

According to the previous paper, *Q* in 1990 is most likely the starting point of the mutation in runoff; therefore, 1970~1989 is set as the base period. In this paper, we focus on the runoff changes in the Weihe River basin in the 2010s, so we set 1990~2009 as the change period PI and 2010~2019 as the change period PII. *ET* increased steadily during the two change periods (Table 4). During the PI period, *P* and *Q* have a similar decline, but the change rate of *Q* is much greater than that of *P*. During the PII period, there was a larger increase in *P* and a certain degree of recovery in *Q*.



Each elasticity coefficient indicates that *Q* is positively correlated with *P* and negatively correlated with *ET* and *n*. The elasticity coefficients (absolute values) of *P* are the largest, with 2.72, 3.06, and 3.17 for each period, reflecting that *Q* is most sensitive to *P* (Table 5). The elasticity coefficients of *n* are −2.01, −2.35, and −2.18 for each period. The elasticity coefficients of *ET* are the smallest: −1.72, −2.06, and −2.17 for each period. The three elasticity coefficients (absolute values) show an increasing trend, indicating that the sensitivity of runoff to *P*, *ET*, and *n* increases at the same time, and *Q* is more susceptible to more drastic changes than the base period, with increased uncertainty and increased chances of flood and drought disasters.


**Table 5.** Elasticity coefficients of each hydrological element in the Weihe River basin at different periods.

#### *4.3. Runoff Change Attribution Identification*

Overall, *n* contributed the most to the variation of *Q*, followed by *P*, and, lastly, *ET*.

During PI, all factors had a decreasing effect on *Q*. An increase of 0.29 in *n* led to a decrease of 13.92 mm in *Q* with a contribution of 66.11% (Figure 4) and was the main cause. A decrease of 22.79 mm in *P* led to a decrease of 6.18 mm in *Q* with a contribution of 29.36% and was the secondary factor. An increase of 10.31 mm in *ET*, resulting in a decrease in *Q* by 0.95 mm with a contribution of 4.53%, was the least influential factor.

**Figure 4.** Contribution of various factors to runoff changes in the Weihe River basin.

During PII, the trend shift in *P* increased by 41.65 mm, resulting in an increase in *Q* of 11.80 mm, with a contribution of −95.56%, offsetting part of the decrease in *Q*. An increase in *n* of 0.46, resulting in a decrease in *Q* of 22.19 mm with a contribution of 179.65% (Figure 4), and was the main cause. An increase in *ET* of 19.01 mm, resulting in a decrease in *Q* of 1.96 mm with a contribution of 15.91%, was the least influential factor on *Q*.

The above shows that changes in each factor cause different degrees of runoff changes over time. *ET* and *n* are negatively correlated with *Q*, and *P* is positively correlated with *Q*. Among them, *n* contributes the most to *Q* changes, and *P* also has a great impact on *Q*. *ET*, because of its own small amount of change, contributes the least to *Q* changes.

From Table 6, we can see that there is a slight difference between the actual runoff depth variation and the calculated runoff depth variation, and the difference between the simulated and actual values in PI and PII periods are less than 0.3 mm, with a relative error of no more than 2%, and the simulated results are very close to the actual values.


**Table 6.** Identification of attribution of runoff changes in the Weihe River basin.

#### **5. Discussion**

The results obtained in this paper, where elevated *n* is the main cause of the sharp decrease in *Q*, are consistent with previous studies [10,13,19], but the contribution of *P* to the change in *Q* is significantly different from previous studies. In this paper, we conclude that *P* increases during the PII period and contributes to an increase in *Q*. Zuo et al., conclude that climate change (*P* and *ET*) contributes 29% to 65% to the decrease in runoff at each hydrological station in the Weihe River basin [10]. Bai et al., conclude that the combination of both climate change and human activities leads to a significant decrease in runoff in the Weihe River main stream [20]. Bi et al., and Liu et al., also reached similar conclusions [21,22]. The differences in the above findings are most likely related to the different study periods, with *P* elevated in the 2010s compared with the 2000s and 1990s, and an increasing effect on *Q*. The former study period was probably in the dry phase of the hydrological cycle, and the decrease in precipitation had a significant decrease in runoff.

*n* is an important factor influencing runoff variation. In this paper, the variation of *n* and its effect on *Q* are analyzed in three periods. To further reflect the changing state of *n*, the meteorological and hydrological data for the whole time period were subjected to a 10-year sliding average, and the corresponding *n* was obtained by back-calculating Equation (4). The change of *n* actually shows the influence of other factors (underlying surface) on runoff change after excluding *P* and *ET*. The increase in *n* indicates that the influence of underlying surface change on *Q* increases, and in this paper, the value of *n* is negatively correlated with *Q*. The increase in its value indicates a stronger effect on *Q* reduction. As can be seen from Figure 5, *n* is in a fluctuating rising state throughout the period, with a significant continuous rising phase after 1995, followed by a gradually declining phase in the 2000s, and finally, a significant rising trend starting around 2008. This indicates that the underlying surface began to change more drastically around 1995 and 2008 than before due to human activities. The overall upward trend in *n* is likely related to increased artificial water withdrawal activity, and a study using reduced natural runoff would likely remove this trend.

**Figure 5.** 10-year sliding average *n* values in the Weihe River basin.

The construction of soil and water conservation projects, the expansion of forest area, the intensification of human water extraction activities, the change of watershed water

storage, and the construction of various water conservancy projects to a certain extent make the process of converting rainfall into runoff more complicated, and the water brought by rainfall is kept in the watershed for a longer period of time, increasing the degree of wetness in the watershed, so the runoff in the watershed will show a decreasing trend for a period of time, which is reflected in the *n* a significant increase. After a period of time, the vegetation coverage and wetness of the watershed reach a certain stage, and *n* is likely to remain more stable or even decline. In 1995, the middle reaches of the Yellow River protection forest creation project began to be implemented, and the first trials were carried out in Shaanxi and other forest areas. The Weihe River basin has since been subsumed under the project of protection forest construction [23,24]. The change in *n* from 1995 until the end of the 2000s is consistent with the pattern of change described above. In 2008, the Shaanxi provincial government launched the comprehensive management project of the Weihe River basin in Shaanxi Province, accelerating the construction of watershed water conservation and ecological projects, forest ecosystem protection and restoration projects, green ecological projects of the Weihe River channel and inter-basin water transfer, etc. From around 2008, *n* again showed an increase, but this state is likely not to last for a long time and is more likely to remain stable or decline in the future.

The impact of future global climate change on the Weihe River basin cannot be accurately simulated, but the increase in temperature and precipitation is recognized and determined at present, which will lead to the change of hydrothermal conditions in the Weihe River basin. The increase in atmospheric temperature and the humidity of the watershed will change the potential evapotranspiration. The increase in precipitation and potential evapotranspiration provides basic conditions for the increase in evaporation. At the same time, the vegetation coverage of the Weihe River basin is also growing rapidly, which makes the change of evaporation more rapid. The increase in precipitation has a positive effect on runoff, but evaporation has the opposite. A very important point is that in the future, Hanjiang to Weihe River Project will add 1.5 billion·m<sup>3</sup> per year of water to the Weihe River. Therefore, there will be many uncertainties in future runoff changes in the Weihe River basin, which require more research on future climate change and human activities.

#### **6. Conclusions**

In this paper, the trend and mutation of hydrometeorological elements in the Weihe River basin from 1970 to 2019 were analyzed using the Mann-Kendall test. the contribution of climate and underlying surface changes to runoff changes were identified by the elasticity coefficient method which is based on the Budyko framework. The main findings are as follows:


an upward trend, which led to a 11.80 mm (accounting for −95.56% of the total runoff change) increase in runoff.

4. In the future, climate change, precipitation, evaporation, and runoff in the Weihe River basin are likely to increase. The increase in vegetation coverage and the interference of human activities will add more uncertainties to the change in the Weihe River runoff. In summary, the runoff of the Weihe River will increase in the future, which requires more comprehensive assessment of climate change and human activities.

**Author Contributions:** Conceptualization, J.X.; methodology, J.X.; validation, T.X.; formal analysis, J.X.; investigation, J.X.; resources, J.X.; data curation, J.X.; writing—original draft preparation, J.X.; writing—review and editing, J.X.; visualization, J.X.; supervision, X.G. and Z.Y.; project administration, J.X.; funding acquisition, Z.Y. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the National Key R&D Program of China (2018YFC1508201), the National Natural Science Foundation of China (51879274) and the China Institute of Water Resources and Hydropower Research (SKL2020ZY03).

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.
