eaec2401-f090-18e7-e053-0937940aa3f2 English nonGeographicDataset non geographic dataset British Geological Survey +44 115 936 3100 Environmental Science Centre, Keyworth NOTTINGHAM NOTTINGHAMSHIRE NG12 5GG United Kingdom enquiries@bgs.ac.uk pointOfContact 2026-04-29 UK GEMINI 2.3 http://data.bgs.ac.uk/id/dataHolding/13607985 Random Forest model output prediction raster for the occurrence of Fe-Mn in the World Oceans 2022-09-28 creation http://data.bgs.ac.uk/id/dataHolding/13607985 The raster provide the output of a machine-learning random forest algorithm modelling the occurrence of ferromanganese (Fe-Mn) crust deposits in the world ocean. This raster constitutes a data-driven approach for mineral prospectivity mapping of Fe-Mn crusts that should be used in conjunction with other expert-driven prospectivity analysis to guide the assessment of Fe-Mn crust coverage in the world ocean and potential mineral exploration. The raster contains values between 0.07 and 0.92. Any values outside of that range (e.g., 0) are outside of the model prediction and should not be displayed. To reproduce data as displayed in the forthcoming associated publication, it is recommended to apply a 'Percent Clip' stretched 'Viridis' colour scheme. Enquiries British Geological Survey 0115 936 3143 0115 936 3276 Environmental Science Centre, Nicker Hill, Keyworth NOTTINGHAM NOTTINGHAMSHIRE NG12 5GG United Kingdom enquiries@bgs.ac.uk distributor Enquiries British Geological Survey 0115 936 3143 0115 936 3276 Environmental Science Centre, Nicker Hill, Keyworth NOTTINGHAM NOTTINGHAMSHIRE NG12 5GG United Kingdom enquiries@bgs.ac.uk originator Enquiries British Geological Survey United Kingdom not available distributor Enquiries British Geological Survey not available originator Enquiries British Geological Survey United Kingdom not available pointOfContact notPlanned https://resources.bgs.ac.uk/images/geonetworkThumbs/eaec2401-f090-18e7-e053-0937940aa3f2.png Geology GEMET - INSPIRE themes, version 1.0 2008-06-01 publication Mineral exploration Ferromanganese Submarine mineral deposits Earth crust BGS Thesaurus of Geosciences 2022 revision UK Location (INSPIRE) Citable Data dataCentre NERC_DDC otherRestrictions restricted The dataset has been formally approved by BGS for access by external clients under a BGS Digital Data Licence, issued via BGS's IPR Section. The dataset must not be provided to, or be accessible by, an external client until the IPR Section confirms that signed licence documentation is in place. Refer to the IPR Section (DigitalLE@bgs.ac.) if further advice is required.. Constraint applies until 2023-10-12 otherRestrictions The copyright of materials derived from the British Geological Survey's work is vested in the Natural Environment Research Council [NERC]. No part of this work may be reproduced or transmitted in any form or by any means, or stored in a retrieval system of any nature, without the prior permission of the copyright holder, via the BGS Intellectual Property Rights Manager. Use by customers of information provided by the BGS, is at the customer's own risk. In view of the disparate sources of information at BGS's disposal, including such material donated to BGS, that BGS accepts in good faith as being accurate, the Natural Environment Research Council (NERC) gives no warranty, expressed or implied, as to the quality or accuracy of the information supplied, or to the information's suitability for any use. NERC/BGS accepts no liability whatever in respect of loss, damage, injury or other occurence however caused. Either: (i) the dataset is made freely available, e.g. via the Internet, for a restricted category of use (e.g. educational use only); or (ii) the dataset has not been formally approved by BGS for access and use by external clients under licence, but its use may be permitted under alternative formal arrangements; or (iii) the dataset contains 3rd party data or information obtained by BGS under terms and conditions that must be consulted in order to determine the permitted usage of the dataset. Refer to the BGS staff member responsible for the creation of the dataset if further advice is required. He / she should be familiar with the composition of the dataset, particularly with regard to 3rd party IPR contained in it, and any resultant use restrictions. This staff member should revert to the IPR Section (ipr@bgs.ac.uk) for advice, should the position not be clear. English geoscientificInformation 2020 2022 raster witheld Enquiries British Geological Survey 0115 936 3143 0115 936 3276 Environmental Science Centre, Nicker Hill, Keyworth NOTTINGHAM NOTTINGHAMSHIRE NG12 5GG United Kingdom enquiries@bgs.ac.uk distributor Enquiries British Geological Survey United Kingdom not available distributor https://webapps.bgs.ac.uk/services/ngdc/accessions/index.html#item176384 WWW:DOWNLOAD-1.0-http--download Data download https://doi.org/10.5285/4c8419b9-5ee4-4db4-b279-18d3ec75c3c4 WWW:LINK-1.0-http--related Citation Information - Digital Object Identifier (DOI) osso, P., Hall, A. (2022). Random Forest model output prediction raster for the occurrence of Fe-Mn in the World Oceans. NERC EDS National Geoscience Data Centre. (Dataset). https://doi.org/10.5285/4c8419b9-5ee4-4db4-b279-18d3ec75c3c4 information nonGeographicDataset non geographic dataset INSPIRE Implementing rules laying down technical arrangements for the interoperability and harmonisation of Geology 2011 publication See the referenced specification false Commission Regulation (EU) No 1089/2010 of 23 November 2010 implementing Directive 2007/2/EC of the European Parliament and of the Council as regards interoperability of spatial data sets and services 2010-12-08 publication See http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2010:323:0011:0102:EN:PDF false The raster dataset correspond to the output of a random-forest algorithm data classification. The model uses a set of known locations for deposit and non-deposit samples to fingerprint the geospatial signature of each class against a range of environmental and geological variables, also referred to as predictors. The multivariate geospatial signature of each pixel covered by all predictors is then used to predict the likelihood of occurrence of the target class. This prediction takes the form of a raster data set with value between 0 and 1 marking the confidence in the prediction of the target class.