Responses of Stream Geomorphic Indices to Piedmont Fault Activity in the Daqingshan Area of China

The Daqingshan piedmont fault spreads along the southern piedmont of the Daqingshan mountains, and it starts in the west from the Zhaojun Tomb in Baotou and extends eastward along the Daqingshan piedmont to the Kuisu area east of Hohhot. This region is an important industrial and economic zone in Inner Mongolia, and this fault is one of the main active faults in the fault system on the northern margin of the Hetao fault-depression zone (He and Ma, 2015; He et al., 2007; Ran et al., 2003; Ma et al., 2000; Research Group of Active Fault System around Ordos Massif, 1988). Extensive research since the 1980s has yielded a substantial understanding of the segmentation, active patterns, geomorphology, and strong quake recurrence of this fault (Wang et al., 2018). More recently, with the emergence of high-resolution satellite remote sensing imaging and geographical information system (GIS) technologies, determining fault activity by extracting and processing the active fault-controlled stream geomorphic indices in the study area has become a principal means of research (He et al., 2019; Dong et al., 2018). Due to the rich presence of geomorphic phenomena in the Daqingshan piedmont, which features multilevel platforms, deeply incised valleys, and river terraces, we aim to, on the basis of previous studies, characterize the tectonic activity of the Daqingshan piedmont fault since the Late Pleistocene by extracting the stream geomorphic indices of the study area from local DEM data with ArcGIS and MATLAB.

The Ordos Block is located in the western part of the North China Craton in the Eurasian continent. A series of faulted basins and folds are present along the periphery, and the neotectonic activity is strong (Fig. 1a). The internal structure of the Ordos Block is stable, but various structures have developed around the block, and several large earthquakes (M8.0–8.5) have occurred in these fault zones and internal faults. Except for the Shanxi fault zone, which began to form in the Pliocene, the fault systems in this area formed in the Oligocene (He et al., 2019; Liang et al., 2019; Xu et al., 2018; Middleton et al., 2016a, b ; Li et al., 2015; Lu et al., 2014; Nie et al., 2010; Deng et al., 1999). The Hetao fault basin is located on the northern margin of the Ordos Block, and its overall trend is approximately east-west. It is kinematically characterized by sinistral shear, and many faults in the basin are normal faults and stepped faults (Fig. 1a). The northern boundary of the basin is the piedmont fault of the Yinshan Range (referred to variably as the Selteng mountains, Wulashan, and Daqingshan)(Fig. 1b). Stratigraphic, geomorphologic and tectonic markers around the fault zone are very clear (Zhang et al., 2019; He et al., 2018, 2017; He and Ma, 2015; Research Group of Active Fault System around Ordos Massif, 1988).

Figure 1.  (a) Ordos perimeter (modified from He et al., 2019; Deng et al., 2007); (b) Daqingshan geology and geomorphology (He and Ma, 2015). 1. Major active faults; 2. Holocene alluvial plain; 3. Late Holocene alluvial fan; 4. Early and Middle Holocene Alluvial fan; 5. Latest Pleistocene piedmont mesa and river terrace; 6. mountainous region 1 400–2 000 m above sea level; 7. mountainous region 2 000–2 400 m above sea level; 8. Quaternary isopach in basin (m). Fault name: F1. Wulashan fault; F2. Daqingshan fault; F3. Helingeer fault.

The ENE-striking Daqingshan piedmont fault is situated on the northern margin of the Ordos Block and represents one of the main active normal faults in the Hetao fault-depression zone (Fig. 1b). The Daqingshan mountains are part of the Inner Mongolian platform uplift of the North China Platform, and its core comprises Precambrian crystalline basement (Inner Mongolia Bureau of Geology and Mineral Resources, 1991). The basic tectonic framework of the Daqingshan area resulted from Yanshanian movement during the Jurassic, and bedrock is exposed to the north of the fault zone. The fault initially formed in the Eocene. Since the Oligocene, the overall fault activity has mainly been characterized by vertical differential movement with an uplifting northern side and subsiding southern side. Normal dip-slip movement has been intense since the Late Quaternary. Faulting has controlled the northern margin of the Hohhot-Baotou depression, wherein the maximum Quaternary thickness is 2 400 m (He and Ma, 2015). The Daqingshan mountains are 2 200–1 600 m high, with the main peak located near the Meidaigou gully at over 2 300 m above sea level (Jiang et al., 2001; Li et al., 1994; Research Group of Active Fault System around Ordos Massif, 1988). Regarding the segmentation of the Daqingshan piedmont fault, Ma et al. (2000) divide the fault into four segments, whereas Jiang et al. (2001) separate it into two segments using Tuzuoqi as the dividing line. In the present study, to further examine the impact of the Daqingshan piedmont fault on the local stream geomorphic indices, we use the fault segmentation proposed by (He et al., 2007) and divide this fault into five segments: the Baotou, Tuyouqi West, Tuzuoqi West, Bikeqi, and Hohhot segments (Fig. 1b).

The Daqingshan piedmont fault has been active since the Late Quaternary. Activity was weak during the Late Pleistocene and increased in the Late Pleistocene. The mountains on the northern side of the Daqingshan piedmont fault have been intermittently uplifted by fault activity. The mountains feature deep valleys and river terraces, and alluvial fans and abrupt mountain fronts are developed in front of the mountain (Fig. 1b). Near the Meidaigou gully, the paleowatershed is less than 7 km from the piedmont fault, suggesting that the Daqingshan mountains are the result of denudation of a south-tilting, north-dipping fault block controlled by the piedmont fault (Ma et al., 1999; Research Group of Active Fault System around Ordos Massif, 1988). The abundance of gullies and streams in the study area (Fig. 2a) provides a good basis for investigating the response of stream geomorphic indices to the activity of the Daqingshan piedmont fault.

Figure 2.  Daqingshan piedmont fault and hydrologic system distribution in the study area (a), AA' topographic profile (b), BB' topographic profile (c).

Topographic swath profiles can be used to statistically calculate the maximum, minimum and mean elevations and the relief degree of the land surface (RDLS) of an area within a given range to quantify the elevation variations among landforms in the area (Liang et al., 2014). The main peak of the Daqingshan Mountains is north of the Meidaigou gully at 2 337.7 m above sea level (Research Group of Active Fault System around Ordos Massif, 1988). The Meidaigou gully is used as a reference position to compile topographic sections in the north-south direction (AA') and in the east-west direction (BB'). In the N-S direction, the elevation of the plain on the north side of the mountain is low. The northern edge of the plain lies at 1 500–1 600 m (Fig. 2b), and the south edge is close to the piedmont fault. The elevation difference rapidly increases towards the south side of the faulted mountain front. There are a large number of peaks above 2 000 m above sea level, and these peaks do not vary much in height and are separated by deep valleys, which indicates that they are part of the same landform. In the E-W direction (Fig. 2b), the terrain in the middle part of the range is undulating and is dissected by the deep valleys. The geomorphological changes near the eastern part are relatively minimal, and the topography is flat. The east-west difference in elevation at the southern margin of the plain (Fig. 2c) reflects the variation in the activity of the Daqingshan fault segments.

In the study, we used 30 m spatial resolution Advanced Spaceborne Thermal Emission and Reflection Global Digital Elevation Model (ASTER GDEM) data (http://www.gscloud.cn/) to extract drainage network using the Hydrology Tools in ArcGIS software (Version 10.5).

Geomorphic landscapes resulting from stream erosion in bedrock channels can reflect the process of river incision (Duvall et al., 2004; Whipple, 2004). In mountainous areas with intense tectonic activity, bedrock channels are often developed, and the erosion rate is mainly determined by the intensity of the river erosion. In this case, the channel erosion rate can be expressed by the river steepness index (Kirby and Whipple, 2012; Hu et al., 2010; Whipple and Tucker, 1999).

where the parameters on the right side of the equation represent the steepness index and concave index of the longitudinal profile of an isostatic channel. Based on Eq. (2), headwater erosion at stream knickpoints is mainly controlled by the drainage area A, although the erosion coefficient K and slope index n also play a critical role in the relationship between the steepness index or channel uplift and erosion (Wang et al., 2016; Zhang et al., 2011; Hu et al., 2010; Whipple et al., 2000).

The hypsometric integral (HI) quantitatively describes the geomorphic evolution in a catchment (Chang et al., 2015). In Davis (1899) proposed the geomorphic erosion cycle theory, which divided geomorphic evolution into three stages: young, mature, and old. Strahler (1952) proposed convex, S-shaped and concave curves according to the basin area-elevation integral curve. There are three variations in shape, and the geomorphological evolution stage is quantified according to the watershed area-elevation integral value (HI): HI < 0.35, the watershed topography is old, and the tectonic activity is weak; HI > 0.6, the watershed topography is young, and the tectonic activity is intense; and 0.6 > HI > 0.35, the watershed is mature, and the tectonic activity is moderate. The HI is often calculated using the simplified formula derived by Pike and Wilson (1971)

where the H_mean, H_max, and H_min are the mean, maximum, and minimum elevations.

Slope and RDLS are among the most frequently used quantitative indices in traditional topographic studies (Scherler et al., 2014). In GIS, slope and flow direction are often extracted from DEM data using a D8 algorithm, that is, through the relationship of a given grid point with the 8 neighbouring points around it (Liang et al., 2018). When the value is greater than 30°, the slope no longer increases with erosion rate (Ouimet et al., 2009). RDLS refers to the difference between the maximum and minimum altitudes in a given area. The slope and RDLS of a mountain area can reflect the intensity of tectonic activity (Whipple and Hancock, 1999; Whipple et al., 1999), and local RDLS values are commonly used in regional studies.

The stream length-gradient index (SL) is an index that quantitatively shows the variation in the gradient of a longitudinal stream profile, and it can serve as an indicator of the intensity of regional tectonic activity (Ji et al., 2011)

where ΔH is the height difference of a river segment, ΔL is the length of a river segment, ΔH/ΔL is the slope of a river segment, and L is the distance from the midpoint to the headwater of a river segment. A larger SL indicates that the tectonic activity in an area is more intense and that the flow rate of streams may be increased by tectonic movement. Hack (1973) drew a Hack surface using the logarithm of the distance from the segment to the headwater of the river as the horizontal axis and elevation as the vertical axis. According to the variation characteristics of the surface, they classified SL into three levels: SL < 200, which signifies a concave-downward shape denoting weak tectonic activity; 400 > SL > 200, which signifies a concave–convex alternation denoting moderate tectonic activity; and SL > 400, which signifies a convex-upward shape denoting strong tectonic activity.

In Fig. 3a, the slope of the mountain near the fault is generally steep, and the overall change in the slope along the fault shows a trend from west to east. While the slope of the piedmont shows the largest change in the slope of the middle section, the slopes of the western and eastern sections gradually decrease. In Fig. 3b, we use the DEM data to extract the topographic relief of the study area. The topographic relief is greatest across the middle portion of the fault, and the topographic relief of the eastern and western sections of the fault is lower. The slope and local topographic relief generally increase from east to west.

Figure 3.  Slope changes (a), terrain fluctuations (b).

We extracted a total of 56 basins along the northern and southern sides of the study area (Fig. 4a) and calculated the SL in each basin. In the north, there are 12 basins (N1–N12), and the SL values are less than 200 (Fig. 4b). In the south, there are 44 basins (S1–S44). Except for the river gradients of S3, S5 and S13, the SL values of all the basins are greater than 200. In Fig. 4b, the variable characteristics of the basins in the south are consistent with the previously present topographic profile (Fig. 2c). To more intuitively represent the SL value of each basin, according to the classification of Hack (1973) (i.e., when SL < 200, the tectonic activity is weak; when 200 < SL < 400, the tectonic activity is moderate; and when SL > 400, the tectonic activity is intense), the basins' values are divided into the 3 categories. Figure 4a shows that, spatially, the fault activity is strong in the middle segment and decreases to the east and west. Additionally, the river gradients generally appear to increase from east to west.

Figure 4.  SL changes (a); SL values (b); area-elevation integral change (c); area-elevation integral values (d).

According to the geomorphic erosion cycle theory, a landscape with a value of HI < 0.35 is old, the terrain is flat, and the slopes are low. For 0.35 < HI < 0.6, the landscape is mature, the terrain is the most rugged and dissected, the river network is well developed, and there is a moderate amount of tectonic activity. For HI > 0.6, the landscape is young, the terrain is flat, and the slope of the ground is low (Xin et al., 2008). The area-elevation integrals were calculated for the 55 basins in the north and south (Figs. 4c, 4d). The area-elevation integral values of the middle section indicate that this section is associated with the high-activity geomorphic evolution stage, which is similar to the conclusions based on the previously presented topographic profile and river gradient results. The terrain is dissected and undulating. The front of the mountain is cut by many "V"-shaped river valleys, and the eastern and western sections feature different HI value ranges. The terrain in these areas is relatively flat, and the slope changes little, unlike in the middle section.

To further quantitatively analyse the characteristics of the regional river geomorphology, the programs ArcGIS and MATLAB were used to automatically extract the longitudinal profiles of rivers from the digital elevation model (DEM) based on a program prepared by Whipple et al. (2007). When extracting the river channels, a 250 m moving window is used to smooth the river channel and eliminate abnormal points. The program yields the steepness index (k_sn), the river length, the river elevation, and other factors (Table 1). To minimize the impact of the collapse on the river channel, the lengths of the selected river channels were no less than 5 km. In Fig. 5a, a total of 78 river channels, R-N1–R-N12 and R-S1–R-S66, were extracted from the basins on the northern and southern sides. The steepness index values of the river channels on the southern side are much higher than those on the northern side. The river channel index values in the middle section of the southern side are the highest, and the values gradually decrease to the eastern and western sections. The steepness index of the river generally shows an increase from east to west (Fig. 5b).

Figure 5.  River channel steepness index (a); characteristics of the steepness index of the river channels (b).

Stream geomorphic indices are typically affected by climate, lithology, and tectonics. Precipitation represents a principal climatic aspect affecting geomorphic indices. The amount of precipitation directly affects the erosion capacity of channels; the lithology of the bedrock directly determines the erosion capacity of streams and channels; and the difference between the tectonic uplift rate and the stream erosion rate in mountainous areas determines the steepness of the stream longitudinal profile. As such, it is necessary to discuss how these factors influence the geomorphic indices in the study area.

Given a constant regional uplift rate, greater precipitation will mean greater stream runoff, a greater stream erosion capacity (expressed as erosion capacity K), and a smaller steepness index (which decreases with an increase in the stream erosion capacity) (Jansen, 2006; Snyder et al., 2000). The Daqingshan piedmont fault extends across the Mongolian Plateau, stretching 200 km from Baotou in the west to Hohhot in the east. The study area is a typical plateau continental climate zone. Yang (1998) compared the sporopollen collected across the Diaojiaohaizi profile of Daqingshan with the paleosol series of the neighbouring area and discovered that the Holocene megathermal in the study area occurred from 9 250-4 000 a BP, and the most suitable climate appeared during 6 850-6 330 a BP, when the climate was warmer (the air temperature was approximately 2 ℃ higher) and more humid than in recent years. After 4 000 a BP, the Holocene megathermal in the area ended, and the climate has become drier and colder since then. Ma et al. (2004) also discovered a possible 10 ka climate cycle in the Daqingshan area. In the present study, we collected recent climate data for the area around the Daqingshan Mountains over a period of 30 years (1981-2010) from the China Meteorological Data Sharing Service System (https://data.cma.cn/), and we found that the recent annual air temperature of the area is approximately 7 ℃. The recent rainfall distribution of the study area was obtained using the linear interpolation technique in GIS (Fig. 6). It was revealed that the multiyear mean annual precipitation of the Baotou-Hohhot area is in the range of 270-410 mm, and the sizes and stream grades of the drainage basins agree well with the spatial variation in rainfall. Overall, the Daqingshan area has experienced a continental arid climate since the Holocene. Thus, it is believed that the climate and precipitation in the study area have remained stable during the Holocene, and they are not the main factors affecting the stream geomorphic variation in the Daqingshan area.

Figure 6.  Distribution of modern precipitation in the Daqingshan area.

As the type of rock varies, the erosional resistance of the bedrock of the riverbed also changes, and the degree of erosion can be determined mainly by the degree of rock fracturing; consequently, different types of bedrock may change the erosion process of a river (Molnar et al., 2007; Whipple et al., 2000). Because this study did not conduct detailed measurements of the strength of the rock in the field, we discuss the differences in bedrock based on the age and lithology of the rock and then judge the possible influence of the lithology on the river geomorphic index. The strata exposed in the study area are dominated by Neoarchean metamorphic rocks, Proterozoic metamorphic rocks and Mesozoic and Cenozoic sedimentary rocks. Magmatic activity occurred frequently in this area from the Archean to the Yanshanian Period (Xu et al., 2017). The rock types mainly include volcanic rocks, breccia, gneiss, deep metamorphic rocks, quartzite, limestone, marble, slate, sandstone, conglomerate, mudstone and clastic rocks. The Precambrian metamorphic rocks and the Mesozoic sedimentary strata constitute the main body of the Daqingshan (Fig. 7). The lower part of the Middle–Lower Jurassic Shiguaigou Formation is composed of sandstone, conglomerate and shale, and the upper part is a set of basic and neutral to moderately acidic volcanic rocks; these rocks are mainly distributed in the western section of the Daqingshan. Permian conglomerate, sandstone and mudstone are mainly distributed in the central Daqingshan. The Upper Jurassic Daqingshan Formation is mainly composed of clastic rock coal seams, which are distributed in the central and eastern sections of the Daqingshan. Quaternary unconsolidated sedimentary layers are generally located in the southern part of the Daqingshan.

Figure 7.  Lithology distribution (modified from Xu et al., 2017; Ren et al., 2013).

According to the lithological distribution characteristics, this study divided the bedrock of the Daqingshan piedmont fault-influenced area into six categories: (1) intrusive rocks (Hua Lixi-Indosinian, Proterozoic, and Caledonian-Hua Lixi); (2) granite (Neoarchean, Proterozoic, Hualixi, Indosinian, and Yanshanian); (3) Neoarchean and Proterozoic gneiss, schist, marble, slate, and other clastic rocks; (4) Cambrian–Mesozoic limestone, dolomite, quartzite, conglomerate, sandstone, mudstone and shale; (5) Triassic white sandstone, conglomerate and clastic rock coal seams; (6) Cenozoic clastic rocks and mudstones and Quaternary sediments. In Fig. 8, the steepness index and lithology do not show obvious regularity or correlation. Previous studies have also shown that lithology has little effect on the steepness index (Hu et al., 2010), so lithology can be ruled out as the main factor controlling the changes in the steepness index in the Daqingshan area.

Figure 8.  Lithology and steepness index distribution of the Daqingshan area. (1) Intrusive rocks (Hualixi-Indosinian, Proterozoic, and Caledonian-Hualixi); (2) granite (Neoarchean, Proterozoic, Hualixi, Indosinian, and Yanshanian); (3) Neoarchean and Proterozoic gneiss, schist, marble, slate, and other clastic rocks; (4) Cambrian–Mesozoic limestone, dolomite, quartzite, conglomerate, sandstone, mudstone and shale; (5) Triassic white sandstone, conglomerate and clastic rock coal seams; (6) Cenozoic clastic rocks and mudstones and Quaternary sediments.

Since the Late Pleistocene faulting, the vertical displacement on the Daqingshan fault has been greatest in the middle (with a maximum value of 148.8 m at Meidaigou) and has gradually decreased to the east and west. To the east and west of the fault, the displacement decreases to more than 50 m. Figure 9 shows that the Daqingshan piedmont fault has been active since the Late Quaternary. The slip rate has more than 4 mm/a since its initiation but has been lower in the Holocene. The slip rate of the highly active middle segment remains above 1 mm/a. The data also show that the slip rate is greatest in the middle section and decreases to the east and west. In previous studies on the Daqingshan piedmont fault, more than 20 trenches were excavated before and after historic earthquake study (Fig. 9). In addition to using trenches to identify paleoearthquake events, river cracks and alluvial fans were also used. The paleosols of the leading margin have been studied to identify ancient earthquake events in the fault section. The number of ancient earthquake events and the average recurrence period during the Holocene also showed that the fault activity has been highest in the middle section and has gradually decreased towards the east and west sections (He et al., 2015; Ran et al., 2003). The differences in the vertical displacement on the fault, slip rate, paleoearthquake activity and other aspects indicate that the fault activity is greatest in the middle section and gradually decreases to the east and west. These trends are consistent with the changes in the local appearance index that was previously obtained. Zhou and Li (2017) used GPS data to study the deformation of the surrounding crust of the Ordos Block. They found that in the northern margin of the Ordos Block, the Hetao Basin has been continuously stretched since 2011, and the entire region is in a state of sinistral shear. This is consistent with the regional tectonic stress state estimated on the basis of geological research. Therefore, since the Late Quaternary, the activity of the Daqingshan piedmont fault has been the main factor controlling the change in the river geomorphology index in the study area.

Figure 9.  Paleoearthquakes and slip rates (modified from He et al., 2007; Ran et al., 2003). Paleoearthquake data: He et al. (2015); Yan et al. (2003).

Based on Shuttle Radar Topography Mission (SRTM) 30-m spatial resolution DEM data, GIS technology was used to extract a variety of geomorphological indices for the study area, and the characteristics of the local appearance index were found to spatially reflect the tectonic activity characteristics of the Daqingshan area. The following conclusions are drawn:

(1) Variations in tectonic activity along the Daqingshan piedmont fault are the main factor affecting the river landform change.

(2) The activity of the Daqingshan piedmont fault has been continuous since the Quaternary. The activity of the middle section from Tuyouqi to Bikeqi is the highest, and the activity decreases towards the eastern and western sections of the fault. The overall activity of the fault gradually increases from east to west.

This work was supported by a research grant from the Institute of Crustal Dynamics, China Earthquake Administration (No. ZDJ2019-21) and the National Natural Science Foundation of China (Nos. 41872227, 41602221). We thank the anonymous reviewers and the editor for their help in improving the manuscript. The final publication is available at Springer via https://doi.org/10.1007/s12583-020-1321-y.