4.1.1. Temporal Change

Since September 2010, the ecological waters have been transferred from the reservoir to Qingtu Lake from September to October each year [40]. The surface area of the water body of Qingtu Lake exhibited a dynamic change trend (Figure 3). The lake water body reached the largest surface area from November to the following January, with a maximum area of 13.7 km2. It normally exhibits the lowest area from July to August each year.

**Figure 3.** The change in the surface area of the water body of Qingtu Lake over time.

## 4.1.2. Spatial Change

Due to the similar temporal trend of the Qingtu Lake water surface area each year, only the spatial distributions of Qingtu Lake water area in di fferent months of 2017 are shown in Figure 4. The water surface area of Qingtu Lake changed little between January and February, ranging between 9.2 km<sup>2</sup> and 9.6 km<sup>2</sup> (Figure 4). After April, the water area began to shrink rapidly, and then reached the lowest level (0.8226 km2) in July. From September, the water area began to expand again, since the water transfer to Qingtu Lake was started in September. The water body expanded to the south and north, and then gradually spread to the west, reaching the highest lake level in November and December. The lake water level then almost stabilized from November to the following February.

**Figure 4.** Spatial distribution of the Qingtu Lake water body from January to December 2017. The spatial trend of the lake water body distribution was similar in other years after 2010.

#### *4.2. Variation of Groundwater Level and Soil Water Saturation in Qingtu Lake Wetland Area*

#### 4.2.1. Variation of the Groundwater Level

During the frozen period, the groundwater monitoring data loggers at locations G02, G08, and G10 were taken out from the wells in December 2018, and thus the data between December 2018 and May 2019 were missing. As shown in Figure 5, the groundwater level at each monitoring location showed a significant rise after 25 August, and then gradually stabilized. After November, the groundwater level at G05, which is located farther away from the lake center, first declined, and after January, the water level at G04 also showed a downward trend because of the groundwater recharge surface water. After March, the groundwater level at G04 and G05 rose obviously. It is speculated that this phenomenon was caused by the rise of the soil surface temperature, which made the frozen surface soil

water melt and recharge the groundwater. In winter, the groundwater at G03, G07, and G09 overflowed the surface, and the groundwater level remained relatively stable for a long time. It can also be seen that the direction of the regional groundwater flow is from G04 to the west and north, and from G9 to G08 and G10.

**Figure 5.** Groundwater-level hydrographs at various monitoring points in the Qingtu Lake wetland area from 2018 to 2019.

Long-term groundwater depth monitoring data at well W01 showed that the groundwater depth gradually decreased from July 2010 to December 2016 under the impact of the water transfer in autumn. The infiltrated surface water greatly increased the groundwater level. Within the same year, the overall groundwater depth fluctuated with greater depth from June to August (Figure 6).

**Figure 6.** Long-term groundwater depth at location W01 near the Qingtu Lake wetland.

#### 4.2.2. Variation in Soil Water Saturation

The water contents were converted to saturation according to Equation (5) [41]:

$$\mathbf{S}\_{\mathbf{r}} = \frac{\boldsymbol{\theta}}{\mathbf{n}}\tag{5}$$

where Sr represents the soil water saturation, θ is the soil water content, and n is the porosity. The monitored water contents were normalized by dividing by soil porosity to more clearly show the change in soil water content (Figure 7). The normalized data shown in Figure 7 refer to the soil water saturation. The V02 and V04 locations are close to the center of the lake, and the saturation at V02 and V04 was between 0.10 and 1. The V07, V08, and V09 locations are farther away, and the groundwater overflowed the surface after the water transfer each year. The saturation at V07 was between 0.388 and 1, the saturation at V08 was between 0.155 and 1, and the saturation at V09 was between 0.078 and 1. The V01, V05, V06, and V10 locations are farthest from the lake, and the groundwater could not submerge the surface; the saturation at V01 was between 0.02 and 1, the saturation at V05 was between 0 and 1, the saturation at V06 was between 0 and 0.471, and the saturation at V10 was between 0and0.791.

**Figure 7.** Variation of soil water saturation at various monitoring points.

As is shown in Figure 7, the variation of saturation at each location (except V06) was similar. The saturation of surface soil at each location was low, while the saturation of deep soil at each location was high. The saturation at the V02, V04, V07, V08, and V09 locations, which are closer to the lake, was low from January to March, and increased after March as the surface soil water melted. The conditions described are consistent with the variation characteristics of the groundwater level.

#### *4.3. Variation in Vegetation Coverage around Qingtu Lake Wetland and Mingqin Basin*

The vegetation classification of the entire Minqin Basin was calculated by referring to the vegetation coverage classification criteria in Table 1. Due to the influence of agricultural crops in the central basin [42], the interpreted vegetation classification with satellite data may not reflect the restoration of the vegetation cover area caused by water transfer. Therefore, this study interpreted the vegetation cover both in the entire Mingqin Basin and the natural vegetation within 10 km around Qingtu Lake. By comparing and analyzing the vegetation coverage degree within 10 km around Qingtu Lake and the entire Minqin Basin, we found that the characteristics of vegetation coverage were similar, and an obvious seasonal change trend was shown in the vegetation coverage areas for each category. Thus, we only showed the temporal and spatial vegetation coverage change within a 10 km area around Qingtu Lake in Figure 8. Generally, the lowest vegetation cover area occurred from the former November to the latter February, and the highest vegetation cover area occurred from June to September. However, there were some di fferences in vegetation cover area for each category. For example, the areas of bare rock, bare soil, and water, with the range of 0–10% vegetation coverage, varied little with the seasons, and the area was basically of the same order of magnitude. However, in the range of vegetation coverage above 30%, the vegetation coverage area varied greatly with the seasons, and the order of magnitude was not uniform. Sometimes, a season with an area of 0 may even occur. It is worth noting that, whether considering the entire Minqin Basin or a 10 km area around Qingtu Lake, the vegetation area with vegetation coverage above 70% was abnormal on 15 July 2010 and 18 July 2011. Compared with the data of the same period in other years, the magnitude di fference was large. After analysis, it was found that the remote sensing images of these two days were from Landsat 4-5 TM and there were clouds, so it is speculated that the reason for this finding may be cloud interference in the remote sensing images. Meanwhile, the total vegetation coverage within a 10 km area around Qingtu Lake showed an obvious increasing trend since 2009 (Figure 8). The vegetation coverage in each category had increased, and the e ffect was most obvious in the vegetation coverage between 30 and 50%.

**Figure 8.** Variations over time of the vegetation coverage for di fferent fractional vegetation coverage (FVC) categories in Qingtu Lake and its surrounding area within 10 kilometers. (Note: There is a lack of Landsat series satellite data from 2009 to 2013, among which there are only two data sets from August and September in 2009, four data sets from July, August and September in 2010, one data set from July, 2011, and two data sets from August and September in 2012).

Figure 9 shows the variation of land use from 1970 to 2017. The obvious changes were in the marshes and low coverage of grass of the natural oasis, which was influenced by the lake area. The area of low coverage grass of the natural oasis increased while the sand area decreased, which was consistent with the variation in vegetation coverage around the Qingtu Lake wetland (Figure 8).

**Figure 9.** Variation of land use from 1970 (**a**) to 2017 (**b**).

#### *4.4. Spatial Distribution of Vegetation Coverage in Qingtu Lake Wetland Area*

The temporal change pattern of vegetation coverage is similar for different years, and thus we only show the spatial distribution of vegetation coverage from January to December 2017 in Figure 10. The areas with high FVC are mainly concentrated in locations close to Qingtu Lake and the south side of the study area, and the FVC of these areas varies greatly with the seasons. In the entire study area, the maximum FVC reached 18.6% from January to April. The maximum FVC reached 67.7% in June. The maximum FVC reached 68.7% in July. The vegetation coverage began to decline after September. In October, the FVC was low, and only sporadic areas had an FVC of more than 30%, while other areas had an FVC of less than 20%. After that, vegetation coverage continued to decline. The FVC, except at Qingtu Lake and the south side of the study area, was lower than 15% for a long time.
