Morphology of the Morasko crater field (western Poland): Influences of pre-impact topography, meteoroid impact processes, and post-impact alterations

Geomorphology 295 (2017) 586–597

Contents lists available at ScienceDirect

Geomorphology journal homepage: www.elsevier.com/locate/geomorph

Morphology of the Morasko crater field (western Poland): Influences of pre-impact topography, meteoroid impact processes, and post-impact alterations Wojciech Włodarski a,⁎, Joanna Papis b, Witold Szczuciński a a b

Institute of Geology, Adam Mickiewicz University in Poznan, Bogumiła Krygowskiego 12, 61-680 Poznań, Poland Institute of Civil Engineering, Poznan University of Technology, Piotrowo 5, 60-965 Poznań, Poland

a r t i c l e

i n f o

Article history: Received 21 April 2017 Received in revised form 8 August 2017 Accepted 8 August 2017 Available online 13 August 2017 Keywords: Impact craters Digital terrain modelling Morphometry Morasko crater field

a b s t r a c t Small impact craters (b 1 km) developed in unconsolidated sediments are expected to be relatively common on Earth; however, only a few tens of them have been documented thus far. Among the reasons for this small number of documented craters are the post-impact erosion and sedimentation processes that modify craters and the lack of universal identification criteria to allow the differentiation of impact structures from landforms of other origins. Here, we focus on the well-preserved impact craters on the Morasko Hill push moraine in western Poland. These craters were formed by iron meteoroid impacts in unconsolidated sediments of glacial and fluvial origins ca. 5000–6000 years ago. We provide a new high-resolution topographic model of the crater field to identify the influences of the pre-impact topography, impact processes, and post-impact modifications on the final morphology of the craters. The topographic model obtained from airborne LiDAR data and total station surveying consists of DEMs related to the recent and reconstructed pre-impact topographies. Parameterization of recent topography in terms of slope gradients, slope curvatures, and roughness allowed us to delimit the boundaries of the craters and to calculate their Feret diameters, ellipticities, slope gradients, crater depths, and volumes. The novelty of our study lies in the estimation of the last two parameters based on the reconstructed pre-impact topography and modelled paraboloids related to each crater. The obtained results show that the studied craters are circular, bowl-shaped features displaying different cross-sectional asymmetries that resulted from the interplay between the trajectories of the bombarding projectiles and the topographies of the primary pre-impact glacial and post-glacial landforms. The oblique impacts likely influenced the asymmetric distribution of ejecta during the excavation of the craters and are considered as factors conditioning mass movements during the postimpact modification of the craters. The compilation of the existing data on terrestrial small-impact craters reveals that they are susceptible to post-impact geometry modification (shallowing and widening) and that many craters have depth/diameter ratios lower than are typical for simple impact craters. © 2017 Elsevier B.V. All rights reserved.

1. Introduction The observations of numerous craters on the surfaces of the Moon, Mars, and other bodies indicate that impact cratering is probably one of the most common geological processes in the Solar System. However, the detail studies of this process have been conducted relatively late, as summarised for instance in review works by Melosh (1989) and Osinski and Pierazzo (2013). On Earth, only 190 impact structures have been verified thus far, as listed in the Earth Impact Database (2017). The majority of these structures are large (mostly 1 to 160 km in diameter) and old impact structures in hard rock substrates. However, impacts of extra-terrestrial hypervelocity bodies capable of forming smaller craters are expected to be much more frequent (French, 1998). Hergarten and ⁎ Corresponding author. E-mail addresses: [email protected] (W. Włodarski), [email protected] (J. Papis), [email protected] (W. Szczuciński).

http://dx.doi.org/10.1016/j.geomorph.2017.08.025 0169-555X/© 2017 Elsevier B.V. All rights reserved.

Kenkmann (2015) presented analysis based on the Near Earth Object (NEO) population, the Lunar crater size frequency distribution, and the mean erosion rate on Earth and estimated that 200–300 impact craters are still to be discovered on Earth and that the missing craters should be small. Bland and Artemieva (2006) assessed the frequency of impacts on Earth's land area by iron bodies of several metres in diameter, which form craters of ca. 0.1 km in diameter, to be at least 1/500 years. Thus, one expects small impact craters on Earth to be relatively common. However, only 16 impact structures b 0.3 km in diameter are listed in the mentioned Earth Impact Database (2017). The reasons for this small number of impact structures include erosion and sedimentation processes that mask the crater forms as well as the difficulties in the identification of impact craters from landforms of different origins, particularly in the cases of impacts in unconsolidated sediments (French and Koeberl, 2010). For instance, in areas glaciated during the Pleistocene, at least two types of landforms are known to be similar to

W. Włodarski et al. / Geomorphology 295 (2017) 586–597

rimmed and circular impact craters. Kettle holes result from melting of glacial dead-ice, and depressions originate from thermokarst periglacial processes related to pingo degradation (e.g., Evans, 2009; van Vliet-Lanoë et al., 2016). Small impact craters are bowl-shaped depressions surrounded by uplifted rims and are called simple craters. Simple craters typically have a depth/diameter ratio of 1:5 to 1:7 (Melosh, 1989). The forms of the small impact structures may also be affected by the pre-impact topography (e.g., Kenkmann et al., 2009), modification stage during crater formation, and post-impact processes (e.g., Gnos et al., 2013). The influence of the pre-impact topography on crater geometry has rarely been studied but appears to be significant (Aschauer and Kenkmann, 2017). For instance, the important role of the pre-impact topography was shown in studies of craters on Vesta, where numerous craters formed on the slopes that locally exceeded 40° (Jaumann et al., 2012; Krohn et al., 2014a). It cannot be excluded that this factor, which interacts with impact trajectory and impact energy, can influence not only the shape but also the size of the developing craters (Elbeshausen and Wünnemann, 2011). This may have important implications because the crater size is the key variable used in crater modelling experiments (e.g., Wünnemann and Ivanov, 2003; Bronikowska et al., 2017). Moreover, the detailed recognition of these interactions allows us to improve the understanding of impact processes and also, to some extent, postimpact changes. The spatiotemporal scale and nature of post-impact changes of terrestrial small craters developed in unconsolidated sediments have been discussed in very few works (e.g., Kenkmann et al., 2009). Most studies focused on impact craters that developed in flatlying, hard rock substrates and show the post-impact changes in terms of the lowering of the crater rims, the degradation of internal crater slopes, and the filling and flattening of crater floors by fluvial, aeolian, and mass wasting processes (Grant, 1999; Komatsu et al., 2014). Although, several terrestrial small impact craters have been relatively well studied (e.g., Carancas, Kaali, Whitecourt, Kamil, and Odessa), a detailed morphometric analysis by means of digital terrain modelling has seldom been presented (e.g., Zanetti et al., 2015). Moreover, among the terrestrial small impact craters, only a few were formed in unconsolidated sediments, which are considered the most common target not only on Earth but also on other planets and moons covered by regolith. The Morasko crater field in western Poland, being the object of the present work, is exceptional, as it is relatively well preserved, taking into account its age of ~5 to 6 ka; and it was formed in unconsolidated sediments of glacial and fluvial origins forming a moraine that exhibits a hummocky topography (Stankowski, 2001). Morasko, a contemporary part of the city of Poznań, is the region with the largest known iron meteorite shower in Central Europe (Muszyński et al., 2012a). The first piece of iron meteorite was found in the area in 1914. However, this and the following findings were not considered to be related to the nearby depressions with diameters of several tens of metres until Pokrzywnicki (1964) suggested that the fall of the Morasko iron meteorites could have resulted in the formation of at least eight impact craters. This view has been debated for a long time, and an alternative glacigenic origin of these depressions owing to the melting of dead-ice blocks was also considered (Karczewski, 1976). The craters were formed in soft, mainly glacigenic sediments in an area of complex glacial topography, formed during the retreat of the last ice-sheet from this area at ~ 18,000 years ago. However, later findings proved the age of the crater infills to be ~ 5000–6000 y BP and, as such, much younger than the last glaciation (Tobolski, 1976; Stankowski, 2001, 2008). Moreover, an ejecta blanket covering paleosols of the same age as the oldest crater infill was recently identified around the craters (Szczuciński et al., 2016). Many additional mineralogical, geochemical, and geochronological studies have provided much data to support the impact nature of the crater field (Stankowski, 2008; Muszyński et al., 2012b). Recently, Bronikowska et al. (2017) combined atmospheric entry modelling, π-group scaling of transient crater size, and hydrocode simulations of impact processes

587

in order to model the Morasko strewn field formation. They used the Morasko crater diameters and distribution as ground truth data. According to their modelling, the most likely entry mass of the meteoroid was between 600 and 1100 tons, the velocity range was between 16 and 18 km s−1, and the trajectory angle 30–43°. The meteoroid was likely subjected to atmospheric breakup, resulting in formation of a strewn field. Bronikowska et al. (2017) suggested that the biggest Morasko crater was likely formed by a projectile 1.5 m in diameter with an impact velocity of ~10 km s−1. The previous morphometric studies on the Morasko craters were initiated by Pokrzywnicki (1964), who presented the first documentation of the crater morphologies and provided their basic dimensions. Later, several standard hypsometric maps of the crater field were made, including that presented by Karczewski (1976). However, no detailed high-resolution modelling of the topography or geomorphological analysis has been performed. The objectives of the present study are to provide a new highresolution topographic model of the Morasko crater field where small impact craters formed in unconsolidated sediments and to apply the obtained model in an attempt to identify the influences of pre-impact topography, impact processes, and post-impact modifications on the final morphology of craters. 2. Study area The study area is located in the northern part of the Morasko Hill push moraine (Fig. 1), marking a major retreat phase of the last glaciation, i.e., the so-called Poznań Phase, ca. 18,500 y BP (Kozarski, 1995). The primary topography of this moraine resulted from glaciotectonic deformation produced during the Vistulian and probably the Saalian glaciations (Karczewski, 1976; Stankowski, 2001). Most of the sediments affected by the glaciotectonic deformation are Quaternary glacial tills, sands, and gravels and Neogene clays, silts, and sands. The maximum elevation of the Morasko Hill push moraine is 153.8 m asl, whereas the topography of the sandur plain extending to the SE varies between 85 and 105 m asl. The minor geomorphologic features are ice-marginal landforms, including small-scale ridges and undrained depressions of evorsive and dead-ice melting origin. The small-scale ridges probably originated through the decay of ice-cored moraines (Ewertowski and Rzeszewski, 2006). In the southern part of the Morasko Hill, these landforms are well-developed, elongated, and linear topographic features oriented ENE-WSW. More irregularly shaped and shorter small-scale ridges tend to form a few lobe-shaped belts in the northern part of the Morasko Hill. These landforms are mainly composed of Vistulian glacial sands and gravels. The glacial landforms were subjected to erosion by small intermittent streams during the post-glacial and Holocene periods and formed the currently dry and small valleys. Some of these valleys are interpreted as beaded valleys (Fleisher, 1986), wherein the wide segments alternate with narrow ones. The wide segments are remnants of depressions produced by dead-ice melting, which were partially reworked by the erosion of intermittent streams. The impact craters were formed before 5000 y BP, as indicated by the radiometric 14C dating of the oldest organic strata filling the craters (Stankowski, 2001). This age is also consistent with the palynological data, which indicate the beginning of organic sedimentation during the middle Atlantic period, between ~ 5500 and 5000 y BP (Tobolski, 1976). Moreover, the thermoluminescence dating of the meltingweathering crust produced by the impact processes gives similar ages of 4700–6100 y BP (Stankowski, 2008). The maximum age of the impact, 5000–6400 y BP, is provided by the age of paleosols buried underneath the impact ejecta layer (Szczuciński et al., 2016). 3. Methodology The present study was performed using digital terrain modelling, starting with processing the digital elevation model (DEM) through

588

W. Włodarski et al. / Geomorphology 295 (2017) 586–597

Fig. 1. Geomorphological and geological setting of the study area. Block diagram showing surficial geology of the study area after the Detailed Geological Map of Poland (Chmal, 1990), draped on LiDAR-based DEM (note the north direction). An arrow marks the probable meteoroid impact from the NE direction at a dip angle of 45° (Muszyński et al., 2012a; Bronikowska et al., 2017). Inset map shows position of the study site in northern Europe.

geomorphometric analysis up to the processing of the pre-impact topography. All computations and spatial data analysis were carried out using ArcGIS 10.4 software (ESRI, 2016), SAGA GIS 3.0.0 open source software (Conrad et al., 2015), and Tableau 10.1 data mining software (Tableau Software, Seattle, WA, USA). The raw data were obtained from airborne LiDAR and total station surveys. The airborne LiDAR data were carried out as a part of the national mapping programme for Poland (Information System of the Protection of the Country against extreme hazards). We excluded these data from the primary input data for processing the DEM because the available elevation points from the LiDAR cloud were collected during a vegetation period (July 2011). Taking into account the dense deciduous forest in the study area, only 2.59% percent of the points were capable of penetrating through canopy gaps and thus were reflected from the ground. How many points really reflected from the bare earth surface and how many of them reflected from stumps and fallen trunks of trees are unknown. For these reasons, the LiDAR-based DEM preprocessed by the Delanuay triangulation is very noisy and does not display all the clearly recognizable topographic features. Thus, the ground-based survey elevation points were used as the primary input data for DEM processing. We surveyed 11,636 accurate elevation points using a Leica TS02 total station and GPS-RTK measurements (Fig. 2A). The measured elevation points were well dispersed as an irregular triangular grid with a mean distance between neighbouring points of 1.68 m and a standard deviation of 0.6 m. Distances between neighbour points were adjusted according to natural changes in the slope gradient and slope curvature of the surveyed surface. A higher density of measured points was applied along the natural breaklines to better display these topographic features in the final model. The uncertainty of the input elevation data (in terms of its possible measurement errors, the microscale variability of the surveyed surface, and its vagueness) was analysed on the basis of the geostatistical modelling of the prediction standard errors using the cross-validation test for simple kriging (Fig. 2B). We assumed a spherical semivariance model with no nugget effect for the polynomially detrended data. Although the first three points in the semivariogram show a slightly asymptotic decrease of

the semivariance from ca. 0.443 to ca. 0.05, we excluded the Gaussian model as it is not natural for elevation data and is probably related to the oversampling of input elevation data (Isaaks and Srivastava, 1989). The calculated mean value of the prediction standard error is 0.22 m, with a standard deviation of 0.05 m (Fig. 2B). The dispersion of the elevation points allowed the gridding of elevation data with an optimal spatial resolution of 0.5 × 0.5 m according to the point pattern analysis method (Hengl, 2006). Gridding was performed using the Topo-to-Raster interpolation method (Fig. 2C). To avoid an oversmoothing effect in the resulting model, the calculated uncertainty of the input data was not included as the vertical standard error in the interpolation (Hutchinson, 1989; Hutchinson et al., 2011). The resulting original DEM was denoised using an effective featurepreserving algorithm (Fig. 2D) to avoid oversmoothing the data and to preserve the sharp topographic features, such as scarps and breaklines (Sun et al., 2007; Stevenson et al., 2010). Based on the uncertainty of the input elevation data, we initially assumed critical angles between the neighbouring modelled grid cells in the range of 4.2°–14°, with a mean of 7.4°, to exclude the denoising procedure for higher angles. The optimal critical angle of 9° was adjusted by trial and error to obtain a good match between the sharpness of the natural breaklines observed in the field and the smoothness of the resulting model. The resulting effects are well illustrated by changes in roughness between the original DEM and the DEM after denoising (Fig. 2D). The roughness maps calculated for two DEMs on the basis of focal statistics, using a window with a radius of 1.5 m (similar to the mean distance between the neighbouring input elevation points), display differences between elevation of each grid cell and mean elevation values of its neighbouring cells. Denoising of the DEM is very important for the geomorphometric analyses focused on the modelling of slope gradients, slope curvatures, and roughness. The DEM-based geomorphometric analysis was performed with visual expert-based boundary delineations of the circular features related to all studied impact craters (Florinsky, 2012; Traglia et al., 2014). For this purpose, the DEM was processed primarily to derive slope gradient, aspect, and profile curvature maps using a 3 × 3 grid-cell moving window. Slope gradient (in °), aspect (in the infinite range from 0° to

W. Włodarski et al. / Geomorphology 295 (2017) 586–597

589

Fig. 2. Flowchart of the main stages in processing of a digital elevation model (DEM) for the Morasko crater field. The labels of craters used here are the same as those used in earlier studies (Pokrzywnicki, 1964; Karczewski, 1976; Stankowski, 2008). See the text for additional explanations.

360°), and profile curvature (defined as the rate of change of the slope gradient) were calculated following the second-order polynomial method of Zeverbergen and Thorne (1987). Maps of the convergence index and the vector ruggedness measure were calculated using operators related to moving windows of 10 × 10 grid cells to obtain more smoothed and artefact-free patterns of ridgelines and terrain roughness respectively (Köthe et al., 1996; Sappington et al., 2007; Olaya and Conrad, 2009). For the delineated circular features, we calculated Feret diameters (min-max), ellipticity, slope gradients (mean, standard deviation, and max), depths of craters (min-max and average), and crater volumes. The novelty of our study lies in the estimation of the last two parameters, which were based on the reconstructed pre-impact topography and modelled paraboloids related to each crater. The reconstruction of

the pre-impact topography was performed for the input data, excluding those from the delimited circular features (Bergonse and Reis, 2015). The masked input data were reinterpolated using the Topo-to-Raster interpolation method. The resulting model was mosaicked with the original DEM using fuzzy logic and then denoised to obtain the final artefact-free model of the pre-impact topography. The modelling of the paraboloids was based on the input of the surveyed elevation data, which were masked to the extent of the delimited circular features. Because three of the craters are filled by lakes, their bottoms were assumed as the bottoms of the drilled organic sediments (Stankowski, 2008). For each crater, we fit a paraboloid to the input data and to the bottom part of the crater using a global second-order polynomial approximation, according to a method of fitting a parabola to crater cross sections (Rehfuss et al., 1977; Kenkmann et al., 2009). Statistics

590

W. Włodarski et al. / Geomorphology 295 (2017) 586–597

related to the slopes, depths, and volumes of the craters were calculated using map algebra. Moreover, to calculate the crater depths, we assumed three elevation values for each crater: the min and max elevation values evaluated along the border of the delineated circular features and the mean elevation value related to the point of the pre-impact topography model located above the deepest part of crater. Another important parameter calculated for the studied craters is impact angle, which means the angle between trajectory of projectile and pre-impact surface bombarded by it. The most probable trajectory for crater-forming projectiles from NW to SW was assumed here based on ellipsoidal spatial distribution (strewn field) of numerous iron meteorites recovered in the study area (Muszyński et al., 2012a). The largest meteorites are found nearby the craters in the SW end of the strewn field. For this trajectory, we assumed the impact angle of 45° as the most probable for terrestrial impacts (Shoemaker, 1962). This assumed angle is plausible taking into account modelled atmospheric entry trajectory of 30 to 43° (Bronikowska et al., 2017). During atmospheric passage and fragmentation, the meteoroid slowed down – so the final impact angle may likely be greater than the one upon the atmospheric entry. The calculations of impact angle for each crater were based on the hillshade mapping of pre-impact topography and refer to points located above the deepest part of particular craters. 4. Results The general morphology of craters labelled A, B, C, E, F, and G is well illustrated in the resulting denoised DEM (Fig. 3). The external

boundaries of these landforms were delineated as circular topographic features along ridge lines seen on the convergence index map and partially along the external boundaries of high slope curvature fields displayed on the vector ruggedness measure map (Fig. 3). The analysed ridge lines display local maxima of slope planar curvatures and are topographic expressions of small-scale divides between the internal slopes of the crater and the sloped surfaces of the primary pre-impact glacial and post-glacial landforms. Moreover, craters A, B, and E are bordered by ridge lines of asymmetrical rims that divide steeply inclined and circular in planview internal slopes from external gentle slopes with more irregular planar curvatures. The asymmetrical rims bordering craters A and B are deeply incised by narrow anthropogenic trenches displaying non-natural, irregular longitudinal profiles. The external boundary of the depression labelled D was delineated only approximately with regard to its irregular shape in planview and its shallow bottom. Because this topographic feature is poorly developed, we labelled it an ill-defined crater. The delineated boundaries of the craters were visually examined using morphological cross sections perpendicular to the internal slopes of the craters (see examples in Fig. 4). Typically, the boundaries displayed continuous transitions from the internal slopes of craters to the sloped surfaces of the primary pre-impact topography. The boundaries were defined here as slight changes in slope gradients, paying additional attention to the circularity of the contour lines derived from the DEM. The geometry of the delimited external boundaries shows that the craters are circular when the ratio between maximum and minimum crater diameters range from 1.03 to 1.102 (Table 1). The exception is

Fig. 3. Final DEM and some of its derivative topographic attributes (slope convergence and terrain ruggedness), as well as reconstructed pre-impact topography. DEM and pre-impact topography are contoured at 5-m intervals. White lines mark boundaries of impact craters. Black lines indicate location of morphological cross sections in Fig. 4. The labels of craters used here are the same as those used in earlier studies (Pokrzywnicki, 1964; Karczewski, 1976; Stankowski, 2008).

W. Włodarski et al. / Geomorphology 295 (2017) 586–597

591

Fig. 4. Examples of morphological cross sections displaying topography of DEM, pre-impact surface, and modelled paraboloids. Note the red arrows that mark boundaries of impact craters. See Fig. 3 for locations of cross sections.

ill-defined crater D, for which this ratio differs considerably (1.58). The maximum crater diameters are in the range of 29.5 to 96.4 m, while the maximum crater depths are from 1.6 to 16 m. Calculated values of the minimum, maximum, and average crater depths (without crater D) are very strongly positively correlated with the maximum crater diameters with Pearson's coefficients of 0.84, 0.94, and 0.97 respectively. The anomalous data related to ill-defined crater D substantially deviates from these regression lines. The internal crater slopes show gradients that differ somewhat according to dip direction and slope length. For example, the asymmetry between the opposite internal crater slopes is well illustrated by the morphological cross sections through craters B, E, and G (Fig. 4), and calculated differences between maximum slope gradients for the opposite internal slopes of these craters are 7.8°, 8.64°, and 7.12° respectively. We have compiled the statistics of the slope gradients partitioned into 15° wide aspect bins for craters A, B, C, E, F, and G (see examples in Fig. 5). Based on maximum and upper quartile values of slope gradients, we recognized bimodal distributions for craters B, C, E, and G; and unimodal and trimodal distributions were observed in craters F and A respectively (see examples in Fig. 5). High slope gradients on the NNE-NE internal

crater slopes are dominant, with the exception of crater F. Less frequent modes of high slope gradients are also apparent on the ENE, W-WNW, and SW-SSW slopes. Notably, high slope gradients show significant positive correlations with slope lengths (Fig. 6). This finding applies both to maximum and upper quartile values of slope gradients, where linear regression models fit the data well based on Pearson's goodness-of-fit test. The obtained correlation coefficients range from 0.3 to 0.97 for the upper quartiles of the slope gradients, while the maximum values of slope gradients correlated with slope lengths have correlation coefficients between 0.25 and 0.97. Despite the wide ranges of these coefficients, strong and very strong correlations were obtained for craters C, E, and G. On the other hand, the weak to moderate correlations found in craters A and B are influenced by the relatively higher slope gradients of the NNE-NE internal crater slopes, displaying relatively short lengths. The asymmetry of the internal crater slopes, described in terms of the different slope gradients and slope lengths, is also expressed in the morphology of the lower emerged parts of the slopes near the craterlake shorelines in craters A and B (Figs. 3 and 5). There are geomorphic indicators of mass movements of surficial materials, including lobeshaped patterns of small-scale, narrow, low-angle slope terraces (see

Table 1 Summary of selected morphometric characteristics of impact craters analysed in the study area. Crater

Diameter (max in m)

Ellipticity (a/b)

Slope gradients (max°)

Depth of craters (min–max in m)

Depth of craters (average in m)

Crater volume based on paraboloids (m3)

Crater volume based on DEM (m3)

Impact angle (°)

A B C E F G D

96.46 64.22 39.99 29.59 33.41 36.75 61.52

1.03 1.05 1.03 1.07 1.07 1.10 1.58

27.99 25.86 27.54 20.42 22.36 23.44 8.20

8.24–15.98 7.27–11.61 4.47–6.27 1.40–2.72 2.14–4.14 1.73–5.15 0.74–1.58

12.13 9.38 5.75 2.12 3.33 4.30 1.22

35,213.11 9909.52 1949.45 483.58 802.06 1851.19 35,213.11

39,364.45 12,075.36 2596.66 594.39 675.39 1525.20 1107.65

49.3 44.3 44.4 42.8 44.2 44.3 45.8

592

W. Włodarski et al. / Geomorphology 295 (2017) 586–597

Fig. 5. Slope gradient maps and sets of box plots for craters A, B, and E providing visual representation of changes in slope gradients in each of 15° wide aspect bins. The maps include grey contours at 5-m intervals and white lines delineating the boundaries of impact craters.

also Fig. 2D). In both cases, mass movement indicators occur on the S-SSE internal crater slopes. Despite this occurrence, in crater A, the internal crater slope affected by mass movements is high and occurs on the uphill side of the primary pre-impact landform; whereas in crater B, this applies to the low internal crater slope on the downhill side of the primary landform. Notably, the crater lake in crater A is strongly offset toward the N-NNW internal crater slope. Because of the bowl-shaped cross sections through dry craters E, F, and G, for which complete input data were obtained from their slopes and bottoms, we considered the modelled paraboloids as more probable general shapes of all studied craters than the cone-shaped models. The goodness-of-fit of these paraboloids to elevations of the recent topography (from the denoised DEM) was evaluated on the basis of their root mean square errors (RMSE), which range from 0.36 to 1.1 m, with the

exception of crater F, where the RMSE is 2.08 m. The cross sections through craters show that the modelled paraboloids are slightly below the bottoms of the craters and the lower to middle parts of the crater slopes displayed on the DEM surface (see examples in Fig. 4). Greater discrepancies between paraboloids and recent topography occur where paraboloids are above the upper parts of the crater slopes (Fig. 4). The intersection lines between the modelled paraboloids and the reconstructed pre-impact surface delineate areas, which occupy 71–86% of the areas related to the external boundaries of the craters. An anomalous percentage of 47% was obtained for crater F. In turn, differences in maximum diameters of craters delineated in the DEM (Fig. 3) and craters modelled from the paraboloids (Fig. 4) range from between 5 and 17%, with the exception of crater F, where an anomalous difference of 43% was observed.

W. Włodarski et al. / Geomorphology 295 (2017) 586–597

593

Fig. 6. Scatter plots of maximum slope gradient to slope length in selected craters. Note the color scales for points that indicate the orientation of internal crater slopes.

The crater volumes, determined as the volumes using the modelled paraboloids and the DEM of the reconstructed pre-impact surface, range from 594 to 39,364 m3 (Table 1). Crater volumes calculated using DEMs related to the present-day and reconstructed pre-impact topography reveal slightly larger volumes, except for those of craters F and G. According to the reconstructed pre-impact topography and the most likely oblique impact direction (to the SW at an angle of 45°), we calculated impact angles for craters A, B, C, E, F, and G (Table 1). The calculated values refer to points of the pre-impact topography located above the deepest part of craters. The greatest impact angle of 49° was obtained for crater A. For other craters, we obtained angles in the range of 42° to 44°. A comparison of recent topography displayed on the DEM surface with the possible reconstructed pre-impact topography revealed that craters were randomly superimposed on various primary landforms (Fig. 3). Crater A was superimposed on the slope of a morainic hill. Craters B and C and the ill-defined crater D were superimposed on opposite slopes of neighbouring morainic hills. Craters G, E, and F were superimposed on different parts of a small valley of an intermittent stream, where wide segments alternate with narrow segments. The internal crater slopes differ in gradients, profile curvatures, and aspects in comparison to the sloped surfaces of the primary pre-impact landforms. The steep internal crater slopes predominate on the uphill side of the

primary landforms and the more gentle internal crater slopes usually occur on the downhill side. Notably, the pre-impact topographic features reconstructed for the crater areas do not create separate landforms but seem to represent a continuation of primary landforms without any major breaks in slope (Figs. 3 and 4), and this can also be applied to the external gentle slopes of the asymmetric rims bordering craters A and B. In the case of crater A, the external slope displays a greater convexity than the slope of the neighbouring pre-impact topography. Nevertheless, the observed convexity is in the lower part of the small valley slope, where slight narrow segments alternate with more wide segments of this valley. With respect to such segmentation, the opposite slope of this valley also shows some similar convexity, with the opposite orientation. The external slope of the rim bordering crater B seems to be a part of a convex-concave slope where the slight planar concavity of the upper part of the slope diminishes gradually downslope, and then increasing planar convexity appears at the lower part of the slope. This slope passes southward into a narrow remnant of the gentle opposite slope, incised by the channel of the intermittent stream. Of all the studied craters, only crater E is partially bordered by a narrow low asymmetric rim, which creates a separate landform at the bottom of a wide segment of a small valley (Figs. 2D and 4). The maximum relative height of this rim is ca. 0.1 m, whereas the width of its external slope does not exceed ca. 3 m.

594

W. Włodarski et al. / Geomorphology 295 (2017) 586–597

5. Discussion 5.1. Morasko craters vs small simple impact craters Depressions within the Morasko crater field, although subjected to several thousands of years of landform transformation, share several properties typical of simple impact craters. These properties include their crater shapes and distributions. First, most of the craters are bowl-shaped circular depressions. The ratios of the maximum and minimum diameters of particular craters approach 1 (Table 1). The only illdefined crater, crater D, is more elongated. However, one may speculate that this crater was formed by two or more projectiles, which fragmented shortly before impact (e.g., Shuvalov and Artemieva, 2015). Second, the important characteristic of the Morasko crater field, which adds to the already existing evidence of their very likely meteoritic origin, is the distribution of the craters. Small impact craters often form crater fields (e.g., Kaali, Odessa, Henbury, Macha; Fig. 7) originated by disintegration of parent meteoroid in the atmosphere. As the Morasko meteoroid entered the atmosphere, it was subjected to at least one fragmentation, leading to disintegration of the meteoroid. Numerical modelling experiments (Bronikowska et al., 2017) revealed that the formation of a single central crater and several symmetrical pairs of smaller craters is a likely scenario. Thus, craters C and E, as well as D and F, form pairs of symmetrical craters with respect to the major crater A. The last pair of craters, B and G, reveal some asymmetry in terms of the crater sizes and distances from crater A. However, it is also possible that the twin meteoroid of the one forming crater B was subjected to further fragmentation and formed much smaller forms. One of them is likely preserved as crater G, which could have had a small twin crater nearby in the past. Simple impact craters are considered to have depth/diameter ratios between 1:5 and 1:7 (Melosh, 1989). This ratio range was established for larger craters; and as shown in Fig. 7, many small impact craters are characterized by smaller ratios, which is also the case for most of the Morasko impact craters. The most likely explanation is that small impact craters can be easily affected by post-impact transformations, including erosion and sedimentation. The most obvious process is the filling of craters with sediment, leading to the reduction of the crater depth. However, crater diameter may also be enlarged by mass wasting processes (e.g., landslides, earthflows) on the crater slopes. Both the filling of the crater and increase in the size of the crater lead to lower depth/diameter ratios.

To explain why some craters still have depth/diameter ratios close to the range of 1:5 and 1:7 and some do not, one needs to consider several factors including the age of the craters, climatic zone (e.g., Nakamura et al., 2014), type of the target rock/sediment, and pre-impact topography (e.g., Aschauer and Kenkmann, 2017). Young craters, for instance, Carancas and Sikhote Alin, seem to be very close to the discussed ratio; while much older craters, such as Odessa or many of the Morasko craters, have much lower ratios. The craters formed in hard bedrock also seem to keep this ratio, even in the case of older impacts (e.g., Kamil, Kaali), while craters of comparable ages in unconsolidated sediments are more easily modified (e.g., Morasko). The roles of the target rock type and regional climatic conditions are well exemplified by two craters in the desert, Kamil in Egypt and Wabar in Saudi Arabia. The first crater is much older, but it is very well preserved (Folco et al., 2010); the second crater is considered to have been formed in historical times and is almost completely filled by windblown sands (Shoemaker and Wynn, 1997). Therefore, in these cases, the availability of the sediment subject to wind transport is a key factor. One should also consider the pre-impact topography, as craters formed on flat terrains are much less susceptible to modification processes than those formed on preexisting slopes. 5.2. Post-impact changes of craters The post-impact changes of the studied craters are expressed by the lack of the impact-formed raised crater rims and by the indicators of mass movements on the internal crater slopes. The lack of raised crater rims can be seen in all the craters, with the exception of crater E where the narrow low asymmetric raised crater rim clearly separates from the primary pre-impact landform, represented by the bottom of a wide segment of a small valley (Figs. 2D, 3, and 4). The asymmetric rims bordering craters A and B do not create separate landforms but represent different portions of the pre-impact morainic hills. Therefore, their impact-related origin is not obvious. In the case of crater B, the possibility cannot be excluded that the original shape and size of the rim was modified by ejecta depositions (Figs. 3 and 4). This rim is very similar to the true impact-induced raised crater rims described from Vesta based on the following characteristics: the downhill position with regard to the pre-impact primary topography, its shape, the asymmetry of the internal crater slope bordering the rim in comparison to the opposite internal crater slope on the uphill side of the primary topography in terms of the slope gradients, and the slope lengths (Krohn et al., 2014a).

Fig. 7. A summary of depth and diameter data for selected small (b200 m diameter) terrestrial impact craters. The grey area represents fields with craters of depth/diameter ratios between 1:5 and 1:7, considered as typical for terrestrial simple craters (Melosh, 1989). In some cases, several depths or diameters were provided because of various reference levels used or ellipticity of craters. Note that in most cases crater depth is reduced according to filling by sediments. The data sources: Carankas - Kenkmann et al. (2009), Dalgaranga - Nininger and Huss (1960), Sikhote Alin - Krinov (1971), Whitecourt - Kofman et al. (2010), Kamil - Folco et al. (2010), Ilumetsä - Raukas et al. (2001), Veevers - Bevan et al. (1995), Morasko present study, Kaali - Veski et al. (2001), Wabar - Shoemaker and Wynn (1997), Henbury - Milton (1972), Odessa - Monnig and Brown (1935), Boxhole - Shoemaker et al. (2005), and Macha - Gurov (1996). A and B present data in normal and log scale, respectively.

W. Włodarski et al. / Geomorphology 295 (2017) 586–597

Moreover, the ridgeline of this rim is smooth and difficult to recognize in the field, while the adjacent internal crater slope is strongly affected by mass movements, despite its low relative height and length. The lack of clear raised crater rims can be explained by gravity reworking of the internal crater slopes during the post-impact stages of crater modification. The observed geomorphic indicators of mass movements can be related to earthflows according to the physicalmechanical properties of the involved materials (Rogers and Chung, 2016) and to the gentle to moderate gradients of the affected internal crater slopes (Highland and Bobrowsky, 2008). Geological studies (Karczewski, 1976; Szokaluk et al., 2016) show that the sand-, silt-, and clay-size grains predominate these mass-moved materials filling the craters and that they originated with the impact-induced transformation of the Pleistocene glacial tills and Neogene clays. The primary frictional and cohesive strengths of the target sediments, namely, the tills and clays, could be overcome with regard to the impact-induced fragmentation of the target materials into breccias and their excavation as ejecta from the developing craters (Nolan et al., 1996). It is likely that just after the excavation stage of the crater formation, the gravitydriven collapse of the ejecta occurred through avalanches and slides under dry conditions (Xiao et al., 2013; Krohn et al., 2014b). In turn, earthflows involving plastic or viscous flows of saturated materials could be triggered later by prolonged or extreme rainfalls. During the mid-Holocene period, the regional climate conditions were similar or wetter than those at present (Tobolski, 1976; Pleskot et al., 2017), supporting the above suggested scenario. The circular-shaped headscarps typical of earthflows (Varnes, 1978) could be mollified and are now difficult to recognize (Figs. 2D and 3). Rogers and Chung (2016) suggested 500 years as the possible timescale for the mollification of earthflow features developed within soft cohesive materials. The observed degree of mollification could result from the overprinting of circular-shaped headscarps on the circular-shaped internal crater slopes. Thus, assuming that the planar curvature of the headscarps was greater than that of the internal crater slopes, we consider that the mollification has resulted in only diminishing the planar curvature to the general trend defined by the curvature of the neighbour portions of the slopes that were not affected by earthflows. Although the occurrence of geomorphic indicators for mass movements of surficial materials is restricted to the S-SSE internal slopes of craters A and B (Figs. 2D, 3, and 4), the post-impact changes driven by these and other possible gravity processes undoubtedly influenced the erosion and degradation of the internal slopes and raised crater rims, as is also observed in other craters (e.g., Kenkmann et al., 2009). The degree of the post-impact changes can be indicated by the differences in the sizes and morphologies of the resulting craters and the modelled paraboloids fitted to the input elevation data (Fig. 4). This approach is possible because of the small size and simple morphology of the studied craters; and therefore, there are similarities in the sizes of the modelled paraboloids and transient craters, which can be assumed to be the final stage of the impact processes before the gravity-driven modifications (Turtle et al., 2005). It is assumed that the possible mass movements can lead to an increase in the crater diameter of 10–20% (Kenkmann et al., 2014). In our case, the differences in the maximum diameters between craters and the areas delineated by modelled paraboloids intersecting the reconstructed pre-impact surface range from 5 to 17% (excluding crater F, where the difference is 43%). 5.3. Primary pre-impact topography versus impact and post-impact processes In addition to the circular in-plane shape of the studied craters, their cross sections display asymmetry in terms of their slope gradients and slope lengths (Figs. 4, 5, and 6). The origin of this asymmetry can be attributed to the oblique impacts resulting from the interplay of two factors: i) the probable projectile trajectories directed to the SW at an angle of 45° and ii) the different topographies of the pre-impact glacial and

595

post-glacial landforms bombarded by projectiles. The probable atmospheric entry angle of the projectiles was assumed to be in the range of 30 to 43° based on mathematical modelling (Bronikowska et al., 2017). Change of the angle during atmospheric passage is unknown. Owing to decceleration it was likely higher than the atmospheric angle, so we assumed it to be ca. 45° – a typical value for impacts on Earth. Taking into account that the craters are very close to each other, it is unlikely to expect large differences between impact angles (assuming a plane horizontal surface) of the crater-forming projectiles. In turn, the impacted topography was evaluated in terms of the changes of the slope gradient and slope aspect. Both slope parameters strongly influenced the obliquity of the impacts at a local scale, according to the fixed trajectories of simulated projectiles. Previous studies have revealed the importance of the relative heights of slopes on the crater shapes if the altitude difference is N30% with respect to the projectile diameter (Elbeshausen and Wünnemann, 2011; Krohn et al., 2014a). According to the most likely simulated diameters of the projectiles bombarding the study area, i.e., in the range of 0.1–2 m (Bronikowska et al., 2017) and the sizes of the resulting impact-induced craters, it can be supposed that the ‘30% rule’ can be applied, and therefore, the primary pre-impact topography strongly affected the impact processes and thus the geometry and size of the craters. Most of the calculated angles of oblique impacts are in the range of 42 to 44° and allow crater asymmetry to be considered as the result of the asymmetric distribution of ejecta during the excavation of the craters (Kenkmann et al., 2014). The range of critical angles where the spatial distributions of ejecta lose their radial symmetry is estimated at between 35 and 45°. The preferred downrange ejection might be related to the S-SSE internal crater slopes, which were most affected by mass movements. Moreover, the opposite N-NNE internal crater slopes displaying highly anomalous gradients are not correlated with the slope length and thus can be related to smaller uprange ejections (Figs. 5 and 6). Notably, the high gradient of the uprange internal crater slope together with the neighbouring low ejecta rim has been cited as a diagnostic of oblique impacts (Kenkmann et al., 2014). Numerical modelling shows that differences in the volumes of downrange and uprange ejected materials are related to different ejection efficiencies controlled by the ejection angles and ejection velocities (Anderson et al., 2003). For vertical impacts, the same ejection angles, between 45 and 55°, in all directions can be expected; whereas for oblique impacts, the ejection angles are low downrange and high uprange. Nevertheless, if we suppose a downrange ejection in the uphill direction following the pre-impact primary topography, then the spatial extent of the resulting ejecta can be limited to the internal crater slopes (Krohn et al., 2014a). This case can be attributed to crater A, where the uphill internal crater slope is strongly affected by mass movements, and the external border of this part of the crater does not display any remnants of a raised crater rim (Figs. 2D, 3, and 4). In turn, the high efficiency and spatial extent of the ejection can be expected in the case of a correlation of the downrange and downhill directions. This possibility can be confirmed by the morphology of crater B, where the internal crater slopes display high asymmetries and where the asymmetric rim modified by the impact processes occurs on the downhill and downrange slope of the primary topography (Figs. 3 and 4). However, some caution should be taken in interpreting the spatial distribution of internal crater slopes affected by mass movements in crater A, as they could be simply controlled by high values of slope length and gradient. It is possible that the post-impact changes related to mass movements overprinted on the primary topography of more circular craters can result in greater asymmetries of internal crater slopes (Krohn et al., 2014a). Similar changes in the asymmetry of the final craters produced by post-impact erosion were discussed by Simonds and Kieffer (1993). In our case, earthflows likely influenced the asymmetries of craters A and B and caused offset of the crater-lake position in crater A toward the opposite internal crater slope and the considerable differences in the slope gradients of the opposite internal slopes of crater B (Figs. 3

596

W. Włodarski et al. / Geomorphology 295 (2017) 586–597

and 4). A similar offset of a crater-lake affected by mass movements was described for the Carancas crater (Kenkmann et al., 2009). In both cases, the slopes affected by mass movements were characterized by their upslope and downrange positions. Numerical modelling of crater formations shows that impact angle can influence peak pressure and thus cratering efficiency, defined as the ratio of ejecta excavated during developing the crater to the initial mass of the projectile (Pierazzo and Melosh, 2000; Elbeshausen et al., 2009). Therefore, we can assume that the size of the greatest crater A is not only controlled by the maximum mass of the bombarding projectile but also by the minimum impact obliquity, defined by the angle of impact, which is higher for crater A than for the other craters, as the pre-impact slope of the present crater A was inclined toward the direction of the projectile arrival. 6. Conclusions The following conclusions can be drawn from the results discussed above: • The studied craters are circular, bowl-shaped features, displaying a symmetrical distribution around the largest crater. • The cross section asymmetry, defined for some craters in terms of slope gradients and slope lengths, resulted from the interplay between the trajectory of the bombarding projectiles and the topography of the targeted pre-impact glacial and post-glacial landforms. • The oblique impacts influenced the asymmetric distributions of ejecta during the excavation of the craters, and thus, the spatial distribution of the internal crater slopes, was affected by mass movements and highly anomalous slope gradients. • The modification of most of the internal crater slopes owing to postimpact mass movements caused the true impact-induced crater rims to be destroyed. • Small terrestrial impact craters are susceptible to geometry modification (shallowing and widening), and many of them have depth/diameter ratios lower than is typical for simple impact craters (b1:5). • The factors affecting the degree of crater modification include the age of the crater, type of target rock or sediment, and type of dominant geomorphologic processes as well as pre-impact topography.

Acknowledgements The study was funded by the Polish National Science Centre (NCN), grant no. 2013/09/B/ST10/01666. We kindly thank the Regional Directorate for Environmental Protection in Poznań for their cooperation and permission to work in the protected nature area. The valuable comments of T. Kenkmann, G. Komatsu, and two anonymous reviewers were of great help in improving the manuscript. The authors thank the Editor Richard Marston for his valuable editorial work. References Anderson, J.L.B., Schultz, P.H., Heineck, J.T., 2003. Asymmetry of ejecta flow during oblique impacts using three-dimensional particle image velocimetry. J. Geophys. Res. 108 (E8):5094. http://dx.doi.org/10.1029/2003JE002075. Aschauer, J., Kenkmann, T., 2017. Impact cratering on slopes. Icarus 290:89–95. http:// dx.doi.org/10.1016/j.icarus.2017.02.021. Bergonse, R., Reis, E., 2015. Reconstructing pre-erosion topography using spatial interpolation techniques: a validation-based approach. J. Geogr. Sci. 25, 196–210. Bevan, A.W.R., Shoemaker, E.M., Shoemaker, C.S., 1995. Metallography and thermomechanical treatment of the Veevers (IIAB) crater-forming iron meteorite. Records of the Western Australian Museum, 17, pp. 51–59. Bland, P.A., Artemieva, N.A., 2006. The rate of small impacts on Earth. Meteorit. Planet. Sci. 41:607–631. http://dx.doi.org/10.1111/j.1945-5100.2006.tb00485.x. Bronikowska, M., Artiemieva, N.A., Wünnemann, K., 2017. Reconstruction of the Morasko meteoroid impact - insight from numerical modeling. Meteorit. Planet. Sci. http://dx.doi.org/10.1111/maps.12882. Chmal, R., 1990. Detailed geological map of Poland at scale 1:50 000. Poznań Sheet. PIG-PIB.

Conrad, O., Bechtel, B., Bock, M., Dietrich, H., Fischer, E., Gerlitz, L., Wehberg, J., Wichmann, V., Böhner, J., 2015. System for Automated Geoscientific Analyses (SAGA) v. 2.1.4. Geosci. Model Dev. 8:1991–2007. http://dx.doi.org/10.5194/gmd-8-1991-2015. Earth Impact Database, 2017. http://www.passc.net/EarthImpactDatabase/index.html. Elbeshausen, D., Wünnemann, K., 2011. The effect of target topography and impact angle on crater formation - insight from 3D numerical modelling. Lunar and Planetary Institute Science Conference Abstracts, p. 1778. Elbeshausen, D., Wünnemann, K., Collins, G.S., 2009. Scaling of oblique impacts in frictional targets: implications for crater size and formation mechanisms. Icarus 204: 716–731. http://dx.doi.org/10.1016/j.icarus.2009.07.018. Evans, D.J.A., 2009. Controlled moraines: origins, characteristics and palaeoglaciological implications. Quat. Sci. Rev. 28:183–208. http://dx.doi.org/10.1016/j.quascirev. 2008.10.024. ESRI, 2016. ArcGIS Desktop: Release 10.4. Environmental Systems Research Institute, Redlands CA. Ewertowski, M., Rzeszewski, M., 2006. Using DEM to recognize possible minor stays of Vistulian (Weichselian) ice-sheet margin in the Wielkopolska Lowland. Quaestiones Geograficae 25A, 7–21. Fleisher, P.J., 1986. Dead-ice sinks and moats: environments of stagnant ice deposition. Geology 14:39–42. http://dx.doi.org/10.1130/0091-7613(1986)14b39:DSAMEON2.0.CO;2. Florinsky, I.V., 2012. Digital Terrain Analysis in Soil Science and Geology. 1st ed. Elsevier, Academic Press, Amsterdam, Boston, Heidelberg. Folco, L., Di Martino, M., Barkooky, A.E., D'Orazio, M., Lethy, A., Urbini, S., Nicolosi, L., Hafez, M., Cordier, C., van Ginneken, M., Zeoli, A., Radwan, A.M., Khrepy, S.E., Gabry, M.E., Gomaa, M., Barakat, A.A., Serra, R., Sharkawi, M.E., 2010. Kamil Crater (Egypt): ground truth for small-scale meteorite impacts on Earth. Geology 39:179–182. http://dx.doi.org/10.1130/G31624.1. French, B.M., 1998. Traces of catastrophe: A handbook of shock-metamorphic effects in terrestrial meteorite impact structures. LPI Contribution #954. Lunar and Planetary Institute, Houston, Texas (120 pp). French, B.M., Koeberl, C., 2010. The convincing identification of terrestrial meteorite impact structures: what works, what doesn't, and why. Earth Sci. Rev. 98:123–170. http://dx.doi.org/10.1016/j.earscirev.2009.10.009. Gnos, E., Hofmann, B.A., Halawani, M.A., Tarabulsi, Y., Hakeem, M., Al Shanti, M., Greber, N.D., Holm, S., Alwmark, C., Greenwood, R.C., Ramseyer, K., 2013. The Wabar impact craters, Saudi Arabia, revisited. Meteorit. Planet. Sci. 48:2000–2014. http:// dx.doi.org/10.1111/maps.12218. Grant, J.A., 1999. Evaluating the evolution of process specific degradation signatures around impact craters. Int. J. Impact Eng. 23, 331–340. Gurov, E.P., 1996. The group of Macha craters in western Yakutia. 27th Lunar and Planetary Science Conference, pp. 473–474. Hengl, T., 2006. Finding the right pixel size. Comput. Geosci. 32:1283–1298. http:// dx.doi.org/10.1016/j.cageo.2005.11.008. Hergarten, S., Kenkmann, T., 2015. The number of impact craters on Earth: any room for further discoveries? Earth Planet. Sci. Lett. 425:187–192. http://dx.doi.org/10.1016/ j.epsl.2015.06.009. Highland, L.M., Bobrowsky, P., 2008. The landslide handbook. A guide to understanding landslides. U.S. Geological Survey Circular, 1325 (129 pp). Hutchinson, M.F., 1989. A new method for gridding elevation and stream line data with automatic removal of spurious pits. J. Hydrol. 106:211–232. http://dx.doi.org/ 10.1016/0022-1694(89)90073-5. Hutchinson, M.F., Xu, T., Stein, J.A., 2011. Recent progress in the ANUDEM elevation gridding procedure. In: Hengl, T., Evans, I.S., Wilson, J.P., Gould, M. (Eds.), Geomorphometry 2011. International Society for Geomorphometry, Redlands, California, pp. 19–22. Isaaks, E.H., Srivastava, R.M., 1989. An Introduction to Applied Geostatistics. Oxford University Press, New York. Jaumann, R., Williams, D.A., Buczkowski, D.L., Yingst, R.A., Preusker, F., Hiesinger, H., Schmedemann, N., Kneissl, T., Vincent, J.B., Blewett, D.T., Buratti, B.J., Carsenty, U., Denevi, B.W., De Sanctis, M.C., Garry, W.B., Keller, H.U., Kersten, E., Krohn, K., Li, J.-Y., Marchi, S., Matz, K.D., McCord, T.B., McSween, H.Y., Mest, S.C., Mittlefehldt, D.W., Mottola, S., Nathues, A., Neukum, G., O'Brien, D.P., Pieters, C.M., Prettyman, T.H., Raymond, C.A., Roatsch, T., Russell, C.T., Schenk, P., Schmidt, B.E., Scholten, F., Stephan, K., Sykes, M.V., Tricarico, P., Wagner, R., Zuber, M.T., Sierks, H., 2012. Vesta's shape and morphology. Science 336:687–690. http://dx.doi.org/10.1126/science.1219122. Karczewski, A., 1976. Morphology and lithology of closed depression area located on the northern slope of Morasko Hill near Poznań. In: Hurnik, H. (Ed.), Meteorite Morasko and the Region of Its Fall. Seria Astronomia vol. 2. Adam Mickiewicz University Press, Poznań, pp. 7–20. Kenkmann, T., Artemieva, N.A., Wünnemann, K., Poelchau, M.H., Elbeshausen, D., Nuñéz del Prado, H., 2009. The Carancas meteorite impact crater, Peru: geologic surveying and modeling of crater formation and atmospheric passage. Meteorit. Planet. Sci. 44:985–1000. http://dx.doi.org/10.1111/j.1945-5100.2009.tb00783.x. Kenkmann, T., Poelchau, M.H., Wulf, G., 2014. Structural geology of impact craters. J. Struct. Geol. 62:156–182. http://dx.doi.org/10.1016/j.jsg.2014.01.015. Kofman, R.S., Herd, C.D., Froese, D.G., 2010. The Whitecourt meteorite impact crater, Alberta, Canada. Meteorit. Planet. Sci. 45:1429–1445. http://dx.doi.org/10.1111/ j.1945-5100.2010.01118.x. Komatsu, G., Senthil Kumar, P., Goto, K., Sekine, Y., Giri, C., Matsui, T., 2014. Drainage systems of Lonar Crater, India: contributions to Lonar Lake hydrology and crater degradation. Planet. Space Sci. 95:45–55. http://dx.doi.org/10.1016/j.pss.2013.05.011. Köthe, R., Gehrt, E., Böhner, J., 1996. Automatische Reliefanalyse für geowissenschaftliche Anwendungen-derzeitiger Stand und Weiterentwicklungen des Programms SARA. Arbeitshefte Geologie 1, 31–37. Kozarski, S., 1995. Deglacjacja północno-zachodniej Polski: warunki środowiska i transformacja geosystemu (ok. 20 ka - 10 ka BP). Continuo, Wrocław.

W. Włodarski et al. / Geomorphology 295 (2017) 586–597 Krinov, E.L., 1971. New studies of the Sikhote-Alin iron meteorite shower. Meteoritics 6, 127–138. Krohn, K., Jaumann, R., Elbeshausen, D., Kneissl, T., Schmedemann, N., Wagner, R., Voigt, J., Otto, K., Matz, K.D., Preusker, F., Roatsch, T., Stephan, K., Raymond, C.A., Russell, C.T., 2014a. Asymmetric craters on Vesta: impact on sloping surfaces. Planet. Space Sci. 103:36–56. http://dx.doi.org/10.1016/j.pss.2014.04.011. Krohn, K., Jaumann, R., Otto, K., Hoogenboom, T., Wagner, R., Buczkowski, D.L., Garry, B., Williams, D.A., Yingst, R.A., Scully, J., De Sanctis, M.C., Kneissl, T., Schmedemann, N., Kersten, E., Stephan, K., Matz, K.D., Pieters, C.M., Preusker, F., Roatsch, T., Schenk, P., Russell, C.T., Raymond, C.A., 2014b. Mass movement on Vesta at steep scarps and crater rims. Icarus 244:120–132. http://dx.doi.org/10.1016/j.icarus.2014.03.013. Melosh, H.J., 1989. Impact Cratering: A Geologic Process. Oxford University Press, New York, NY. Milton, D.J., 1972. Structural geology of the Henbury meteorite craters, Northern Territory, Australia. U.S. Geol. Surv. Prof. Pap. 599-C, C1–C17. Monnig, O.E., Brown, R., 1935. The Odessa, Texas, meteorite crater. Pop. Astron. 43, 34–37. Muszyński, A., Kryza, R., Karwowski, Ł., Pilski, A.S., Muszyńska, J. (Eds.), 2012a. Morasko. The Largest Iron Meteorite Shower in Central Europe. Bogucki Wydawnictwo Naukowe. Muszyński, A., Stankowski, W., Pilski, A.S., Kryza, R., Nowak, M., 2012b. Iron meteorite shower Morasko, Przełazy, Jankowo Dolne. In: Muszyński, et al. (Eds.), Morasko. The Largest Iron Meteorite Shower in Central Europe. Bogucki Wydawnictwo Naukowe, pp. 27–35. Nakamura, A., Yokoyama, Y., Sekine, Y., Goto, K., Komatsu, G., Senthil Kumar, P., Matsuzaki, H., Kaneoka, I., Matsui, T., 2014. Formation and geomorphologic history of the Lonar impact crater deduced from in situ cosmogenic 10Be and 26Al. Geochem. Geophys. Geosyst. 15:3190–3197. http://dx.doi.org/10.1002/2014GC005376. Nininger, H.H., Huss, G.I., 1960. The unique meteorite crater at Dalaranga, western Australia. Mineral. Mag. 32, 619–639. Nolan, M.C., Asphaug, E., Melosh, H.J., Greenberg, R., 1996. Impact craters on asteroids: does strength or gravity control their size? Icarus 124:359–371. http://dx.doi.org/ 10.1006/icar.1996.0214. Olaya, V., Conrad, O., 2009. Geomorphometry in SAGA. In: Hengl, T., Reuter, H. (Eds.), Geomorphometry - Concepts, Software, Applications. Developments in Soil Science vol. 33. Elsevier, pp. 293–308. Osinski, G.R., Pierazzo, E. (Eds.), 2013. Impact Cratering: Processes and Products. WileyBlackwell. Pierazzo, E., Melosh, H.J., 2000. Understanding oblique impacts from experiments, observations, and modeling. Earth Planet. Sci. Lett. 28:141–167. http://dx.doi.org/10.1146/ annurev.earth.28.1.141. Pleskot, K., Tjallingii, R., Makohonienko, M., Nowaczyk, N., Szczuciński, W., 2017. Holocene paleohydrological reconstruction of Lake Strzeszyńskie (western Poland) and its implications for the central European climatic transition zone. J. Paleolimnol. (in review). Pokrzywnicki, J., 1964. Meteorites of Poland (in Polish with English summary). Studia Geologica Polonica 15, 9–140. Raukas, A., Tiirmaa, R., Kaup, E., Kimmel, K., 2001. The age of the Ilumetsa meteorite craters in southeast Estonia. Meteorit. Planet. Sci. 36:1507–1514. http://dx.doi.org/ 10.1111/j.1945-5100.2001.tb01842.x. Rehfuss, D.E., Michael, D., Anselmo, J.C., Kincheloe, N.K., Wolfe, S.A., 1977. Hyper-ballistic transport models of Copernican ejecta. Proceedings of the Lunar Science Conference, 8, pp. 3375–3388. Rogers, J.D., Chung, J.W., 2016. Mapping earthflows and earthflow complexes using topographic indicators. Eng. Geol. 208:206–213. http://dx.doi.org/10.1016/ j.enggeo.2016.04.025. Sappington, J.M., Longshore, K.M., Thompson, D.B., 2007. Quantifying landscape ruggedness for animal habitat analysis: a case study using bighorn sheep in the Mojave Desert. J. Wildl. Manag. 71:1419–1426. http://dx.doi.org/10.2193/2005-723. Shoemaker, E.M., 1962. Interpretation of lunar craters. In: Kopal, Z. (Ed.), Physics and Astronomy of the Moon. Academic Press, San Diego, pp. 283–359. Shoemaker, E.M., Wynn, J.C., 1997. Geology if the Wabar meteorite craters, Saudi Arabia. 28th Lunar and Planetary Science Conference, Abstract No. 1660.

597

Shoemaker, E.M., MacDonald, F.A., Shoemaker, C.S., 2005. Geology of five small Australian impact craters. Aust. J. Earth Sci. 52:529–544. http://dx.doi.org/10.1080/ 08120090500180921. Shuvalov, V.V., Artemieva, N.A., 2015. Craters made by severely fragmented asteroids. 46th Lunar and Planetary Science Conference, p. 1442. Simonds, C.H., Kieffer, S.W., 1993. Impact and volcanism – a momentum scaling law for erosion. J. Geophys. Res. 98:14321. http://dx.doi.org/10.1029/93JB00704. Stankowski, W.T.J., 2001. The geology and morphology of the natural reserve “Meteoryt Morasko”. Planet. Space Sci. 49:749–753. http://dx.doi.org/10.1016/S00320633(01)00011-3. Stankowski, W., 2008. Meteoryt Morasko osobliwość obszaru Poznania - Morasko Meteorite, A Curiosity of Poznań Region. 2nd edition. Wydawnictwo Naukowe UAM. Stevenson, J.A., Sun, X.F., Mitchell, N.C., 2010. Despeckling SRTM and other topographic data with a denoising algorithm. Geomorphology 114:238–252. http://dx.doi.org/ 10.1016/j.geomorph.2009.07.006. Sun, X., Rosin, P.L., Martin, R.R., Langbein, F.C., 2007. Fast and effective feature preserving mesh denoising. IEEE Trans. Vis. Comput. Graph. 13:925–938. http://dx.doi.org/ 10.1109/TVCG.2007.1065. Szczuciński, W., Szokaluk, M., Bronikowska, M., Jagodziński, R., Muszyński, A., Wünnemann, K., 2016. Identification and dating of small impact crater ejecta deposits, case of Morasko craters, Poland. 32nd IAS International Meeting of Sedimentology (2016) (Marrakech, Morocco). Szokaluk, M., Muszyński, A., Jagodziński, R., Szczuciński, W., 2016. Properties of ejecta blanket deposits surrounding Morasko meteorite impact craters (Poland). 79th Annual Meeting of the Meteoritical Society August 7–12, 2016. Meteoritics and Planetary Science vol. 51, p. 6516. Tobolski, K., 1976. Palynological investigations of bottom sediments in closed depressions. In: Hurnik, H. (Ed.), Meteorite Morasko and the Region of Its Fall. Seria Astronomia vol. 2. Adam Mickiewicz University Press, Poznań, pp. 21–26. Traglia, F.D., Morelli, S., Casagli, N., Monroy, V.H.G., 2014. Semi-automatic delimitation of volcanic edifice boundaries: validation and application to the cinder cones of the Tancitaro-Nueva Italia region (Michoacán-Guanajuato Volcanic Field, Mexico). Geomorphology 219:152–160. http://dx.doi.org/10.1016/j.geomorph.2014.05.002. Turtle, E.P., Pierazzo, E., Collins, G.S., Osinski, G.R., Melosh, H.J., Morgan, J.V., Reimold, W.U., 2005. Impact structures: what does crater diameter mean? In: Kenkmann, T., Hörz, F., Deutsch, A. (Eds.), Large Meteorite Impacts III. Geological Society of America Special Paper, 384. Geological Society of America, Boulder:pp. 1–24 http://dx.doi.org/ 10.1130/0-8137-2384-1.1 Van Vliet-Lanoë, B., Brulhet, J., Combes, P., Duvail, C., Ego, F., Baize, S., Cojan, I., 2016. Quaternary thermokarst and thermal erosion features in northern France: origin and palaeoenvironments. Boreas http://dx.doi.org/10.1111/bor.12221. Varnes, D.L., 1978. Slope movement types and processes. In: Schuster, R.L., Krizek, R.J. (Eds.), Landslides: Analysis and Control. Transportation Research Board Special Report vol. 176, pp. 11–33. Veski, S., Heinsalu, A., Kirsimäe, K., Poska, A., Saarse, L., 2001. Ecological catastrophe in connection with the impact of the Kaali Meteorite about 800–400 BC on the island of Saaremaa, Estonia. Meteorit. Planet. Sci. 36:1367–1376. http://dx.doi.org/ 10.1111/j.1945-5100.2001.tb01830.x. Wünnemann, K., Ivanov, B.A., 2003. Numerical modelling of the impact crater depth– diameter dependence in an acoustically fluidized target. Planet. Space Sci. 51: 831–845. http://dx.doi.org/10.1016/j.pss.2003.08.001. Xiao, Z., Zeng, Z., Ding, N., Molaro, J., 2013. Mass wasting features on the Moon - how active is the lunar surface? Earth Planet. Sci. Lett. 376:1–11. http://dx.doi.org/10.1016/ j.epsl.2013.06.015. Zanetti, M., Wilk, J., Kukko, A., Kaartinen, H., Kohv, M., Jõeleht, A., Välja, R., Paavel, K., Kriiska, A., Plado, J., Losiak, A., Wisniowski, T., Huber, M., Zhu, M.H., 2015. The structure of the Kaali Impact Crater (Estonia) based on 3D laser scanning, electroresistivity tomography, and iSale hydrocode modelling. Bridging the Gap III: Impact Cratering in Nature, Experiments, and Modeling, Held 21–26 September, 2015 at University of Freiburg, Germany. LPI Contribution No. 1861, p. 1103. Zeverbergen, L.W., Thorne, C.R., 1987. Quantitative analysis of land surface topography. Earth Surf. Process. Landf. 12:47–56. http://dx.doi.org/10.1002/esp.3290120107.