Identification of maars and similar volcanic landforms in the West Eifel Volcanic Field through image processing of DTM data: efficiency of different methods depending on preservation state

The West Eifel Volcanic Field comprises 98 maars, tuff rings, and scoria rings of volcanoes younger than 700 ka. Digital Terrain Models (DTMs) allow to automatically measure morphologic parameters of volcanic edifices such as slope angles, diameters, elevations, floor, and slope surface areas. Based on their morphological characteristics, we subdivided the West Eifel volcanoes into five morphometric groups which reflect different stages of erosion. Group I, II, and IV comprise clear ring-shaped structures. The difference between these groups is that a tephra ring is well preserved in Group I, partially preserved in Group II and absent in Group IV. The original shapes of Group III maars have been lost more substantially than in Groups I, II, or IV, but they nevertheless retain a negative shape (a depression) and have characteristic channel systems, which can be used as search criteria. Maar-diatremes of Group V are eroded down to their feeder pipes and form hills. In order to locate potential volcanic depressions that are likely to be maar volcanoes, we defined common search criteria such as circular negative landforms or particular drainage system patterns for all groups except the least well-preserved Group V. These criteria were taken as the basis for further processing of the DTM data. The first processing step consisted of constructing a residual relief calculated as the difference between a filtered (smoothed) topographic surface and the original DTM data. This identifies local topographic features. We propose a method for regulating the degree of smoothing which is based on filtering of local maxima according to their distance from a surface constructed from local minima. The previously defined search criteria for Groups I to IV such as specific ranges of curvature, slope, circularity, density of the drainage network were then applied to the residual relief in order to extract maar shapes. Not all criteria work equally well for all morphological groups. Combinations of multiple search criteria therefore yield the best results and efficiently identify most known maars. They also separate some probable new, hitherto unrecognized maars from a large number of other local depressions. We also compared the erosional state of maars to their absolute ages. Published estimates of erosion rates for maars in the French Massif Central suggest a general trend of erosion rates decreasing with time elapsed since eruption. However, this cannot explain the strongly varying ages for maars of the same morphometric group (i.e., similar preservation state) in the West Eifel Volcanic Field. The spatial distribution of the morphometric groups shows some regularity. For example, strongly eroded maars are concentrated in the Gerolstein area (where maar density is highest), whereas most well-preserved maars are located east of the Eifel North-South Depression (ENSD). Most maars affected by fluvial erosion lie near the Kyll and Kleine Kyll streams. These observations suggest differential recent uplift of the West Eifel Volcanic Field, with stronger uplift occurring west of the ENSD.