A machine learning approach to geochemical mapping

A peer revied paper in Journal of Geochemical Exploration. Geochemical maps provide invaluable evidence to guide decisions on issues of mineral exploration, agriculture, and environmental health. However, the high cost of chemical analysis means that the ground sampling density will always be limited. Traditionally, geochemical maps have been produced through the interpolation of measured element concentrations between sample sites using models based on the spatial autocorrelation of data (e.g. semivariogram models for ordinary kriging). In their simplest form such models fail to consider potentially useful auxiliary information about the region and the accuracy of the maps may suffer as a result. In contrast, this study uses quantile regression forests (an elaboration of random forest) to investigate the potential of high resolution auxiliary information alone to support the generation of accurate and interpretable geochemical maps. This paper presents a summary of the performance of quantile regression forests in predicting element concentrations, loss on ignition and pH in the soils of south west England using high resolution remote sensing and geophysical survey data.



Through stratified 10-fold cross validation we find the accuracy of quantile regression forests in predicting soil geochemistry in south west England to be a general improvement over that offered by ordinary kriging. Concentrations of immobile elements whose distributions are most tightly controlled by bedrock lithology are predicted with the greatest accuracy (e.g. Al with a cross-validated R2 of 0.79), while concentrations of more mobile elements prove harder to predict. In addition to providing a high level of prediction accuracy, models built on high resolution auxiliary variables allow for informative, process based, interpretations to be made. In conclusion, this study has highlighted the ability to map and understand the surface environment with greater accuracy and detail than previously possible by combining information from multiple datasets. As the quality and coverage of remote sensing and geophysical surveys continue to improve, machine learning methods will provide a means to interpret the otherwise-uninterpretable.



Website:

https://nora.nerc.ac.uk/id/eprint/513707/

Simple

Date (Creation)
2016-08-01
Date (Publication)
2016-08-01
Date (Revision)
Citation identifier
https://www.mica-project.eu/ / MICA_B2-3

MICA WP3 2017-09-25T10:33:00 Record extracted from Batch 2 spreadsheet

Other citation details

No additional information provided about the dataset
Purpose

to improve geochemical mapping techniques
Status
Completed
Point of contact
Organisation name Individual name Electronic mail address Role

British Geological Survey

 		Resource provider
Maintenance and update frequency
Not planned
dataCentre

  • MICA

MICA ontology (TemporalScheme)

  • Recent (includes data from 2006 onwards)

MICA ontology (DataScheme)

  • General descriptive information / Paper
Keywords

  • Requirement for data generation: Voluntary
Keywords

  • Method of data or information generation: Academic research

GEMET Concepts

  • Exploration

GEMET Concepts

  • Geochemical

GEMET Concepts

  • Geology

GEMET Concepts

  • Geophysical

GEMET Concepts

  • Gravity information

GEMET Concepts

  • Mineralogical

GEMET Concepts

  • Minerals

GEMET Concepts

  • Sciences dealing with the composition, structure, origin of the Earth's rocks

GEMET Concepts

  • Soils

INSPIRE commodities

  • Gold

INSPIRE commodities

  • Silver

INSPIRE commodities

  • Platinum

INSPIRE commodities

  • Palladium

INSPIRE commodities

  • Other platinum group metals (iridium)

INSPIRE commodities

  • Other platinum group metals (osmium)

INSPIRE commodities

  • Other platinum group metals (rhodium)

INSPIRE commodities

  • Other platinum group metals (ruthenium)

INSPIRE commodities

  • Aluminium

INSPIRE commodities

  • Copper

INSPIRE commodities

  • Lead

INSPIRE commodities

  • Tin

INSPIRE commodities

  • Zinc

INSPIRE commodities

  • Chromium

INSPIRE commodities

  • Cobalt

INSPIRE commodities

  • Iron

INSPIRE commodities

  • Manganese

INSPIRE commodities

  • Molybdenum

INSPIRE commodities

  • Nickel

INSPIRE commodities

  • Niobium

INSPIRE commodities

  • Tungsten

INSPIRE commodities

  • Vanadium

INSPIRE commodities

  • Antimony

INSPIRE commodities

  • Beryllium

INSPIRE commodities

  • Bismuth

INSPIRE commodities

  • Cadmium

INSPIRE commodities

  • Cesium

INSPIRE commodities

  • Gallium

INSPIRE commodities

  • Germanium

INSPIRE commodities

  • Hafnium

INSPIRE commodities

  • Indium

INSPIRE commodities

  • Lithium

INSPIRE commodities

  • Magnesium

INSPIRE commodities

  • Mercury

INSPIRE commodities

  • Rare earths (undifferentiated)

INSPIRE commodities

  • Rhenium

INSPIRE commodities

  • Rubidium

INSPIRE commodities

  • Selenium

INSPIRE commodities

  • Tantalum

INSPIRE commodities

  • Tellurium

INSPIRE commodities

  • Titanium

INSPIRE commodities

  • Zirconium

MICA ontology (DomainScheme)

  • Detailed geochemistry

MICA ontology (DomainScheme)

  • Geostatistical estimates

MICA ontology (DomainScheme)

  • 2D predictive mapping

MICA ontology (DomainScheme)

  • Regional geochemistry
Access constraints
Other restrictions
Use constraints
Other restrictions
Language
English
Topic category
  • Geoscientific information
  • Economy
Description

Soth West Region, United Kingdom
Geographic identifier
Region within a country (please provide more details)
Hierarchy level
Non geographic dataset
Other

Individual item (e.g. a one-off academic paper or single website)

Conformance result

Title

Data uncertainty
Date
Explanation

Are any uncertainty measures provided (e.g. standard errors, confidence intervals, etc.)?
Pass
Yes

Conformance result

Title

Quality assurance procedures
Date
Explanation

Are quality assurance procedures described?
Pass
Yes

Conformance result

Title

Information generation methods
Date
Explanation

Are data or information generation methods formally described?
Pass
Yes

Conformance result

Title

Record review
Date
Other citation details

Reviewed by: British Geological Survey
Explanation

Record validation
Pass
No

Conformance result

Title
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
Date (Publication)
2010-12-08
Statement

Information provided on possible sources of digitisation error, etc.

Metadata

File identifier
0de044a6-13b5-4675-9ff2-863ca84b5815 XML
Metadata language
English
Character set
MD_CharacterSetCode_utf8
Hierarchy level
Non geographic dataset
Date stamp
2024-11-07
Metadata standard name

ISO19115
Metadata standard version

2003/Cor.1:2006
Metadata author
Organisation name Individual name Electronic mail address Role

British Geological Survey

enquiries@bgs.ac.uk

Point of contact
 
 

Overviews

overview

Keywords

MICA


Provided by

Share on social sites

Access to the catalogue