Today we will search for minerals using data openly available on the Google Earth Engine (GEE) platform, using geological modeling and subsequent support vector classification to predict gold-bearing areas from the built geological model. We will need ALOS terrain with a resolution of 30 m, Sentinel-1 SAR radar images with a resolution of 10 m and Sentinel-2 10 m optical images (for visualization only). The accuracy of the classifier turned out to be 97.77% and, most importantly, the result meets the expectations of the geologist - the found areas are actually very promising.
An area for detailed study is marked with a red and white ball.
Introduction
, , , , , . , . , , , - , . , , , .
— , , , , . , , , . ( ) , , , . , . , , . , .
. ( ) ( ) . , , ( ). , ( 5 Sentinel-1 , ), , , . , , (PostgreSQL/PostGIS , ).
, . , , . , , . 3 / (3 ppm) 5 /, 4 12 . , , , .
, . , , :
, , , . , ( ) .
:
, . :
— , «». (SVC) ( ) . "" , , 300-500 ( ). :
, , . — , - , - . , , . - , , . , , :
, , — , .
, , :
, () — , , , . :
() . , , , (. ) . , ( , ). , , ( -, , ) ( , ) .
, , , . ( ), — . , , . , , - .
- HOWTO: Multi-scale 3D Gravity and Magnetic Inversion modeling; https://www.linkedin.com/pulse/brief-explanation-3d-density-depth-model-construction-pechnikov/
- Gold grade of epithermal gold ore at Lamuntet, Brang Rea, West Sumbawa District, West Nusa Tenggara Province, Indonesia. AIP Conference Proceedings 1855, 050002 (2017); https://doi.org/10.1063/1.4985516 Rika Ernawati, Arifudin Idrus, and Himawan TBMP
- Sentinel-1 SAR GRD: C-band Synthetic Aperture Radar Ground Range Detected, log scaling; https://developers.google.com/earth-engine/datasets/catalog/COPERNICUS_S1_GRD
- Sentinel-2 MSI: MultiSpectral Instrument, Level-2A; https://developers.google.com/earth-engine/datasets/catalog/COPERNICUS_S2_SR
- ALOS DSM: Global 30m; https://developers.google.com/earth-engine/datasets/catalog/JAXA_ALOS_AW3D30_V3_2
- C-Support Vector Classification sklearn.svm.SVC; https://scikit-learn.org/stable/modules/generated/sklearn.svm.SVC.html