Spatial variations prediction in carbonate porosity using artificial neural network: Subis Limestones, Sarawak, Malaysia
The estimation and modeling of carbonate porosity is of increasing interest in different aspects of geology. Several models have been developed to visualize the pore network systems of carbonate rocks. However, no modeling tools have been designed to predict changes in pore system resulting from dis...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Published: |
Springer Nature
2019
|
| Online Access: | https://www.scopus.com/inward/record.uri?eid=2-s2.0-85108113955&doi=10.1007%2f978-3-030-01440-7_44&partnerID=40&md5=bba797a117cea58f17fce9d057e445c5 http://eprints.utp.edu.my/23485/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The estimation and modeling of carbonate porosity is of increasing interest in different aspects of geology. Several models have been developed to visualize the pore network systems of carbonate rocks. However, no modeling tools have been designed to predict changes in pore system resulting from dissolution. Therefore, this paper introduced an algorithm for predicting spatial variations in pore network. Carbonate outcropped samples representing different facies from Subis limestone, Sarawak, of Miocene age were used for this study. Continuous imagery along a 10 cm rock chip was conducted using Micro Computer Tomography (CT) scan imagery. The Artificial Neural Network (ANN) predictive code receives images which were read as a matrix. The images were processed using the Image Analysis, coded before use as a training and input data set for ANN. The ANN produced a predicted image with the same properties (such as bits, scalar or raster etc.) as the input images and at the same interval. The predicted image was compared to the original one to estimate the prediction accuracy. The method proved to give good results in terms of the predicted images accuracy. The method can be applied to study the dissolution phenomenon in carbonates as well as siliciclastic rocks to predict spatial variations and development in a pore network system. © Springer Nature Switzerland AG 2019. |
|---|