Quantitative mapping and predictive modeling of Mn nodules' distribution from hydroacoustic and optical AUV data linked by random forests machine learning

Abstract

Abstract. In this study, high-resolution bathymetric multibeam and optical image data, both obtained within the Belgian manganese (Mn) nodule mining license area by the autonomous underwater vehicle (AUV) Abyss, were combined in order to create a predictive random forests (RF) machine learning model. AUV bathymetry reveals small-scale terrain variations, allowing slope estimations and calculation of bathymetric derivatives such as slope, curvature, and ruggedness. Optical AUV imagery provides quantitative information regarding the distribution (number and median size) of Mn nodules. Within the area considered in this study, Mn nodules show a heterogeneous and spatially clustered pattern, and their number per square meter is negatively correlated with their median size. A prediction of the number of Mn nodules was achieved by combining information derived from the acoustic and optical data using a RF model. This model was tuned by examining the influence of the training set size, the number of growing trees (ntree), and the number of predictor variables to be randomly selected at each node (mtry) on the RF prediction accuracy. The use of larger training data sets with higher ntree and mtry values increases the accuracy. To estimate the Mn-nodule abundance, these predictions were linked to ground-truth data acquired by box coring. Linking optical and hydroacoustic data revealed a nonlinear relationship between the Mn-nodule distribution and topographic characteristics. This highlights the importance of a detailed terrain reconstruction for a predictive modeling of Mn-nodule abundance. In addition, this study underlines the necessity of a sufficient spatial distribution of the optical data to provide reliable modeling input for the RF.

Ecology of a polymetallic nodule occurrence gradient: Implications for deep‐sea mining

“…The CoMoNoD algorithm calculates the size of each nodule (i.e., seafloor exposed area) detected in an image, enabling the calculation of descriptive nodule statistics. Note that it is currently not possible to directly relate the image‐based assessments of seabed nodule cover with those made by direct sampling methods (Schoening et al ; Gazis et al ). Megafauna specimens were identified to the lowest taxonomic level possible, and their physical dimension was measured, using BIIGLE 2.0 (Langenkämper et al ).…”

Deep-ocean polymetallic nodules as a resource for critical materials

“…4). In addition to bathymetric information, such systems provide acoustic imagery (backscatter) that can be used to interpret the geological conditions on the ocean floor, including the separation of flat, sediment-covered areas that contain nodules from areas devoid of them 45,[79][80][81][82][83][84] . After areas of interest (which can cover several thousand square kilometres) have been identified, a detailed bathymetric survey is conducted using autonomous underwater vehicles or deep-towed acoustic systems, such as side-scan sonar.…”

“…After areas of interest (which can cover several thousand square kilometres) have been identified, a detailed bathymetric survey is conducted using autonomous underwater vehicles or deep-towed acoustic systems, such as side-scan sonar. Underwater systems survey at tens of metres up to about 100 m above the ocean floor, thus providing higher resolution (less than a few metres) bathymetric surveys than achievable from vessel-mounted systems (50-100 m) 82,83 . High-resolution acoustic survey are typically accompanied by video mapping using either deep-towed video or photo sledges, remotely operated vehicles or auto nomous underwater vehicles 3 (Fig.…”

“…Data gathered by sample collection and highresolution acoustic surveys are then fed into multivariate (geo)statistical models to predict nodule abundance and metal resources over large areas 5,78,79,83,84 . Such predictive models are based on artificial neural networks or random forests, as well as on classical geostatistics, such as variography and kriging, in order to assess the quantity of the resource 5,83,[85][86][87] . Once the economic potential of a nodule field has been determined, then, if favourable, the operations will move on to mining.…”

Extensive Coverage of Marine Mineral Concretions Revealed in Shallow Shelf Sea Areas

“…In many coastal sea areas, the term ferromanganese concretion is established to underline the high concentration of these elements in shallow-water precipitates. Despite the widespread occurrence of shallow-water concretions, considerably more research effort has been invested in studying the distribution of deep-sea nodules (Gazis et al, 2018;Peukert et al, 2018b;Alevizos et al, unpublished), with only scattered observations of shelf sea and coastal concretions (EMODnet Geology, 2019).…”