What is beneath the “hat-shaped hills” of Denmark?

English translation of Danish article “De hatformige bakkers underlag” in ‘Geologisk Tidsskrift’ 2000:1, Dansk Geologisk Forening, pp.1-2, written by Peder Dam & Johnny Jakobsen, University of Roskilde

The socalled “hat-shaped hills” of Denmark have been the subject of many discussions and theories among Danish geologists throughout the century, ever since their unusual internal structure was discovered by Vilhelm Milthers and Kristian Rørdam in 1900. The hills in question seem rather “normal” on the outside. They are found in plain moraine clay areas as singular round hills of various heights, usually until 20-30 meters above ground level and about 200 meters long. They are often located in series or rows, in places where glacial border lines are expected to have been during the end of the Weichsel glacial period. The hills consist of sand and gravel sedimented in layers just like normal kames of stagnant glacial lake deposits, but to the contrary of these, the layers in the hat-shaped hills seem to be tilted and are often found standing almost vertical, but still with very well-stratified layers. On the top of the hills, there are sometimes to be found a thin layer of moraine clay.

The first hat-shaped hill, which would later on be considered the “type location” of the term, was Gedebjerg near Kundby in NW-Zealand. A large part of the hill has been removed and turned into a gravel pit, which provides a fine profile through the hill of about 100 meters. Two smaller hills are situated right to the south-east of Gedebjerg and they also contain dislocated layers of gravel and sand. In all, the existence of about a dozen of such hat-shaped hills have been established in the Kundby area. Later on, other geologists have found hills of very similar appearance in other parts of Denmark. The largest hills of this specific type is the three Kulsbjerge near Vordingborg in SE-Zealand, while the island of Langeland is practically covered by hat-shaped hills; more than one thousand hills are concentrated on the small, oblong island. Our inquiries directed to geologists and geological institutes in Nothern Europe and America indicates, that phenomena - more or less similar to the Danish hat-shaped hills - are known in the USA and Canada, and especially in Poland and the Baltic states.

The rather silly name for the Danish hills was given to them by Milthers back at the turn of the century, because their shape reminded him of a certain type of ladies hats, which were in fashion at the time. An other and more “scientific” name for the hills is “dislocated kames”, but the problem with this name is that it rules out all but one theory of the construction of the hills - and it is exactly the question of their construction which has given cause for all the theories and investigations of the hills.

The reason for the lively debate of the hat-shaped hills for all these years is that it has been quite a challenge to present a complete explanation of their shape, the vertical strata and the complicated tectonic structure, why the uppermost part of the strata is overtilted, the possible cover and “train” of till, together with the location and grouping of the hills.

Sedimentological facies analysis and tectonical analysis have been made in especially Gedebjerg up to our time. Based on these, it has been tried to explain how the sedimentations originally took place and how the layers later on were dislocated. However, the studies have not been able to determine the making of the hat-shaped hills with certainty, but they have produced enough new knowledge to modify the old theories and create new ideas as well. Both the analysis results and the validity of the different theories have continuously been questioned by fellow geologists.

By performing a new set of investigations, using geoelectric testings for the first time in the matter of hat-shaped hills, we (a group of geology students at the University of Roskilde) tried to establish, what is underneath the hills. This is of interest, as the theories can be divided into two groups; one that says the material came “from above” and was - one way or the other - put on the moraine bed by a glacier in the end of the Weichsel period and then tilted by the same or another glacier. The other group of theories bring up the idea, that the material was originally sedimented by an older glacier and later pushed up “from beneath” by a younger glacier.

The very first and classic theory of the origin of the hills was made by Milthers (1900), who suggested that the sediments were originally deposited in at least 200 meters deep crevices in an ice-marginal environment. The tilting of the strata to almost vertical happened during the melting away of the supporting ice walls. Andersen (1966) argued for a nearshore beach environment of an interglacial sea, where later ice movements should have brought large pieces of frozen sea sediments with them to the Kundby area. The idea of dislocated kames was presented by Rasmussen (1967), where the vertical layers of sediments should origin from glacial lakes in ice which was still moving or the sediments had been pushed over by a later ice.

If one from the first group of theories was correct, one should expect to find moraine clay beneath the hills as well as in the terrain around them. This, however, was not the case. A number of geoelectric measurements, performed by us in the autumn of 1998 and the spring of 1999, clearly established, that while moraine clay was found around and between the hills, the ground right beneath Gedebjerg and a small neighboring hill consisted of sand or gravel until a depth of at least 50 meters. Due to the conditions in the hill and the capacity of our equipment, is wat not possibly to perform any valid measuring any deeper than this.

Profiles of the geoelectrical soil tests mapped from two different angles. The grey colour indicates sand and/or gravel, the black colour indicates moraine clay.

The results will come as a surprise for most geologists with an interest in the matter, as the most popular theories till today have been, that the hills were sedimented right on the thick moraine till bed by one of the very last glaciers in the area. Our geoelectric findings have support by other observations, however. Local farmers have informed us, that subsoil water leak out of Gedebjerg in the early spring quite a few meters above ground level, while lower parts of the plain around the hill is still dry. Also the gravel pit was given up years ago, when they reached permanent wet sand at a height of about four meters above the surrounding terrain. These significant signs of sand/gravel beneath Gedebjerg, was confirmed by our tests.

Both Vilhelm Milthers and his son Keld Milthers performed countings of indicator boulders all over the country back in the first half of the century. Such countings were also performed in two of the hat-shaped hills of Kundby, Gedebjerg and Kirkebjerg. Their findings did not go well with the Milthers’ own idea of the making of the hills, as the stones in Gedebjerg and especially Kirkebjerg appeared to origine from the earlier glaciers in the Weichsel period (with home base in Norway and Sweden), while the rest of the region had a larger percentage of stones from the Baltic Sea, which is considered to be the base of the last ice covering Zealand.

Two theories support the fact of gravel and sand beneath the hills. One was presented by Asger Berthelsen in the 1970’s and the other by Niels Schrøder in 1997. They both build on the idea, that old sedimentations were pushed up in openings of the moraine plain by the pressure of a later ice, which caused the significant dislocation of the layers, but the models differs in the exact mechanism of the event. Berthelsen speaks of a relative simple overturned anticlinal foldstructure, while Schrøder compares the situation with the reaction in salt diapires.

With our investigations, we do not claim to have found the correct theory of the origin of the hat-shaped hills. The geoelectric measurements do not give any indications of support for one of the two “gravel-ground”-theories on behalf of the other. Actually, we cannot even totally reject the “moraine-ground”-theories, but these theories have to be modified to cope with the new knowledge. Also we must underline, that our testings only includes Gedebjerg and the small hill to the west of it. We cannot say anything about, whether similar results would appear if geoelectrical measurements were performed in hat-shaped hills elsewhere in Denmark. But since Gedebjerg is the definition of a “hat-shaped hill”, we do have added a new perspective to the mystery of this Danish phenomenon, which will enter another century unsolved.