Monitoring and Characterization of Surface Deformation after the Closure of Coal Mines Based on Small Baseline Interferometric Synthetic Aperture Radar

3.1 Datasets

Table 2. List of ASAR images for InSAR analysis

The study area was observed in two stages. In the first stage, the spatial distribution and rate of surface deformation were estimated based on ten L0-level images (descending pass; C-band) of the study area taken by Envisat’s advanced synthetic-aperture radar (ASAR) from August 29, 2006 to August 3, 2010 (Table 2). The topographic phase contributions were removed by the digital elevation model (DEM) data (spatial resolution: 30m) of Shuttle Radar Topography Mission (SRTM). The orbital accuracy of the ASAR images was improved based on the Doris orbit data released by the European Space Agency (ESA).

In the second stage, 134 Sentinel-1A/B images (descending pass; TOPS mode; C-band) captured on 99 days from March 2015 to December 2018 were analyzed (Table 3). The research area is located on the edge of these images. The ESA’s precise orbit ephemerides (POD) were adopted as the orbit data.

5.1 Spatial correlation between settlement and uplift

The comparison between Figures 4 and 5 shows that, the settlement areas of the first stage basically correspond to the uplift areas of the second stage, although the settlement centers does not coincide with the uplift centers.

Figure 6 shows the profile of settlement centers A-E in the south (sandstone pattern) and that of the corresponding uplift areas (black curve). The profile line is about 20km from southwest to northeast. The data that are missing due to decorrelation in the first stage were neglected. It can be seen that the uplift curve was much gentler than the settlement curve, carrying no abrupt changes as the settlement curve. There is a bulge in the uplift curve of the second stage. The peak settlements of the five basins were 152, 154, 169, 126, and 172mm, and the corresponding uplifts were 63, 90, 58, 52, and 61mm, respectively. The uplifts are about 41%, 58%, 34%, 48%, and 35% of the peak settlements, respectively, with an average of 42%.

6.png

Figure 6. The profiles of settlement centers in the south and corresponding uplift areas

In the settlement profile, there is a -20 to -50mm (-5 to -12mm/year) settlement between the basins, in addition to the bowl-shaped settlements in the basins. The uplifts at the corresponding positions exceeded the settlements. These results demonstrate that uplift and settlement are roughly correlated, rather than proportionate to each other.

Mining operations can only proceed in dry space. During coalmining groundwater is continuously pumped out from the roadways and surrounding rocks, resulting in a decline of groundwater level in and around the mine. As a result, a depression cone is formed in the regional distribution of groundwater. After the coalmine is closed, the pumping immediately comes to an end. Then, the groundwater in the mine is recharged. The rising groundwater level leads to a rising pore pressure and the expansion of underground clay, which in turn bring surface uplift. The uplift is often proportional to the increment of groundwater level. In the settlement basins, the groundwater level rise faster than the surroundings. That is why the surface above the basins exhibits uplift motion.

5.2 Nonlinear evolution of uplift

To quantify the surface uplift in the second stage, the time series of surface deformation at all uplift centers were plotted (Figure 7). In the left subgraphs, the red dots are the observed values of uplift, and the blue lines are fitted by the Richards equation. In the right subgraphs, the blue lines are the fitted curves of deformation rate.

As shown in Figure 7, the surface deformation was nonlinear in the second stage. The deformation at most uplift centers, except for I, could be divided into two phases: rapid uplift phase (the uplift rate increased to the peak before slowing down) and uplift mitigation phase (the uplift was slow and gradually dropped to zero, i.e. the deformation stabilized). Uplift center I was different from other uplift centers, due to the early closure of the mine (December 2014).

To fully understand the deformation features, the Richards equation [29] was employed to fit the deformation curve of each uplift center. This equation models the nonlinear biological growth: slow growth, fast growth, slow growth, and equilibrium. The data fitting helps to unveil the evolution law of deformation time series. The Richards equation can be expressed as:

7a.png

7b.png

7c.png

Figure 7. Time series of deformation at each uplift center and fitted curves of deformation rate

Table 6. Fitting results of each uplift center in the second stage

where, y(t) is the predicted deformation at time t; a is the progressive limit of deformation, i.e. final deformation; b is the initial deformation coefficient; k is deformation rate coefficient; m is the shape parameter of the deformation curve. The deformation rate at time t can be obtained through first-order derivation:

$v(t)=\frac{a b k}{1-m}\left(1-b e^{-k t}\right)^{\frac{m}{1-m}} e^{-k t}$          (2)

The fitting results (Figure 7) show that the surface deformation clearly sped up and then slowed down in the second stage. Table 6 lists the maximum deformation rates fitted at I-N. It is clear that the deformation process cannot be described reasonably with only one average value (Vervoort, 2016).

The nonlinear surface deformation is mainly attributable to the change of groundwater level. Once the mine is closed, the groundwater continues to flow towards the mine pit. As the drainage is terminated, the recharge amount far outnumbers the outward penetration. Thus, the groundwater level will rise at a faster speed. However, the higher the water level, the more the diversion channels. The outflow volume will continue to increase, and the water level will grow at an increasingly slow rate. The relationship between groundwater and surface water will become more and more complex. The groundwater level will reach the highest level, when the amount of pit water equals the amount of infiltration. The resulting dynamic equilibrium is different from the natural state and the normal production stage.

5.3 Influence of geological conditions

The above analysis suggests that in the south of the coalfield, each settlement center basically corresponds to an uplift (of course, this is not applicable to all mines). Take Shuanggou mine and its surroundings for example. In the first stage, the peak settlement of -148mm occurred on the northern surface of the mine, while almost no deformation took place in the second stage.

As shown in Figure 8, there was also a 60-80mm uplift step along Mansihe fault, i.e. the northern boundary of Shuanggou mine. Manshihe is an east-west trending fault with a drop of 60-80m, rising in the north and falling in the south. The fault does not contain any water. Shuanggou mine is bordered in the east by Panlongshan fault (drop: 140m), which rises in the west and falls in the east, in the west by Wangmushan fault (drop: 350-550m), and in the south by Hongshan mine, where sufficient protective coal pillars are reserved.

8.png

Figure 8. The uplift step in the north of Shuanggou mine

Shuanggou mine has simple hydrogeological conditions. There are four groups of aquifers in the mine: the aquifer group of Ordovician fissure limestone, the aquifer group of middle Carboniferous Xujiazhuang fissure limestone, the aquifer group of sandstones on coal seam roof and floor and thin fissure limestone, and the aquifer group of fissure sandstone above coal-bearing stratum. The first two groups have high water resistance, thanks to the even distribution of clay shale, mudstone and other water resisting strata. Most of the latter two groups are covered by upper Quaternary yellow laterite, whose underdeveloped fissures and weak water bearing property limit the infiltration of atmospheric precipitation.

Due to the above geological conditions, the groundwater was recharged slowly after Shuanggou mine was closed. As a result, the groundwater level rose by a limited margin. This explains why the surface deformation was unobvious in the mine, and why uplift step appeared along the Manshihe fault in the north. Of course, the surface of this mine will be lifted with further recharge of groundwater. It would be interesting to investigate the surface deformation in this area.

5.4 Relationship between surface uplift and underground mining location

Zhangzhao mine and the area in the northwest of Fengshui mine were analyzed to clarify the possible relationship between surface uplift and underground mining location.

Zhangzhao mine started operation in 1987 and was closed at the end of 2014. There are four minable coal seams: 0.57m-thick coal seam #3, 0.67m-thick coal seam #5, 0.62m-thick coal seam #10-1, and 1.01m-thick coal seam #10-2.

The northwest of Fengshui mine was included into the mine in 2004. In the south part of this area stands a petrochemical plant. Thus, the coal under this part was not mined. The mining operations in the northwest of Fengshui mine lasted from 2006 to October 2015. There are two minable coal seams: the 0.6-1.0m-thick coal seam #3, and the 1.01m-thick coal seam #5.

Figure 9 presents the distribution of mining locations in Zhangzhao mine and the northwest of Fengshui mine. In the first stage, no settlement was observed in either region. In the second stage, both regions were uplifted. The surface uplift of Zhanghao mine was between 10 and 57mm. The areas with greater-than-40mm uplift mainly concentrate in the south, center, and east of the mine. The largest uplift of 57mm appeared in area A. The other areas with greater-than-50mm uplift include areas B (50mm), C (51mm), D (51mm), and E (51mm). Among them, area C is the only area with mining operations.

The surface uplift in the northwest of Fengshui mine fell between 6 and 54mm. The areas with greater-than-40mm uplift concentrate in the southeast. The maximum uplift (54mm) appeared in the area of the petrochemical plant, though no mining operation was performed here. Among the areas with mining operations, the maximum settlement occurred in area B (44mm). Hence, mining location does not strictly correspond to surface uplift, and surface uplift might occur in non-mining areas.

The above phenomena demonstrate that coalmining not only affects the groundwater level, but also alter the original form of groundwater dynamic field, due to fault structure, mining fissures on the floor, and various manmade well holes. The mining operations may even cause the groundwater to move across strata, resulting in abnormal change of groundwater level. This indirectly leads to the uplift of the surface.

9.png