Lidar Processing for Defining Sinkhole Characteristics under Dense Forest Cover: A Case Study in the Dinaric Mountains

Abstract

ABSTRACT:The traditional approach for defining sinkholes characteristics is based on topographic maps and air photographs with derived digital terrain models. This method is sometimes not accurate, requiring costly, time consuming and potentially dangerous fieldwork. Investigations have shown that airborne scanning laser data (lidar) is useful in detection of karst depressions due to the high density of ground points that can be obtained. This is especially important under dense forest canopy, where classical photogrammetric methods do not allow ground points to be measured. The objective of this work was to map and determine geomorphometric characteristics of a large number of sinkholes located in a diverse karst terrain under a dense forest tree-canopy using lidar data. We tested an algorithm described in previous literature which uses only information from the DTM. It is based on water flow simulations on a surface (DTM) and incorporates four phases: (i) watershed delineation, (ii) confining of sinkholes, (iii) confining of higher rank sinkholes and (iv) extraction of non-karstic sinkholes. Sinkholes were confined by effluent level with cells below the effluent level designated as part of the sinkhole. In the third step sinkholes were ranked according to their location and size -first rank sinkholes are the smallest and are located within a larger sinkhole. Results are that the sinkhole fraction of 1st, 2nd, 3rd, 4th and 5th rank in the study area was 3.25%, 4.26%, 5.68%, 3.65% and 3.14%, respectively. Sinkhole distribution shows a peculiar directionality in their spatial distribution, which seems to be significantly towards a northwest -southeast direction. It was not possible to compare results with ground-truth data due to very low accessibility, nevertheless a statistical and visual assessment of the results shows that lidar is a very effective technique to model sinkholes under dense canopy.

Geomorphological Dating of Pleistocene Conglomerates in Central Slovenia Based on Spatial Analyses of Dolines Using LiDAR and Ground Penetrating Radar

“…New methods for automated doline detection and delineation as well as numerous tests of their reliability are lately subject to considerable expansion [5,7,8,28,29]. Additionally, more and more precise input data (LiDAR) for such researches has recently been made available, but issues such as understanding the concept of surface karstification and the pitfalls of new methods should be considered when trying to obtain results of the same quality as those expected from high-resolution data.…”

“…The increasingly available high-resolution LiDAR (Light Detection and Ranging) method for obtaining elevation data and creating digital elevation models (DEMs) is still in early stages of ability and reliability testing compared to traditional topographic and aerial surveying methods in modern karst geomorphology [5]. Researches applying high-resolution DEMs are mainly focused on automation of dolines detection and delineation [5][6][7][8], rather than on interpretations of the acquired geomorphological data [4,9].…”

Detecting Terrain Stoniness From Airborne Laser Scanning Data †

“…The vertical height difference (depth) was considerably larger than in in [23,24]. The high density point cloud and a carefully designed multi-step process results in quantitative analysis of sinkholes in [25], unlike in our study, where the stoniness likelihood of a binary classifier is the only output.…”

“…The vertical height differences are at the same range, 0.5-1.5 m in our study and in [23,24], though. A similar study of [25] uses higher density LiDAR data with ρ = 30 m −2 to detect karst depressions of size 26 m and more. The vertical height difference (depth) was considerably larger than in in [23,24].…”

Evaluating susceptibility of karst dolines (sinkholes) for collapse in Sango, Tennessee, USA

“…Over the last two decades, Geographic Information Systems (GIS) has become an important tool and has been increasingly used for mapping karst and creating karst databases (Gao and Zhou, 2008; Gao and Alexander, 2003; Gao et al, 2002, Siska and Kemmerly, 2008; Kobal et al, 2014). Gutierrez et al (2014) calls karst inventories ‘the most important step in hazard analysis.’ For the last decade, the study of dolines using remote sensing was also extended to the planet Mars where the existence of dolines in evaporite karst has drawn an increasing amount of attention from the research community (Baioni et al, 2015).…”