An Advanced Calibration Method for Image Analysis in Laboratory-Scale Seawater Intrusion Problems

Abstract

Image analysis is a useful tool for visualising flow through laboratory-scale aquifers but existing methods of converting image light intensity to concentration can be labour intensive and time consuming. The new approach proposed in this study utilises the Random Forest machine learning technique to build a calibration model to replace the requirement for unique calibrations of each test aquifer. Calibration images from a previous experimental study were used to train the Random Forest model and the output was compared to the results from a high resolution pixel-wise methodology. The Random Forest model provided a trade-off in accuracy with increased efficiency and reduced sensitivity to image desynchronisation when compared to the pixel-wise method. The reduced accuracy was attributed in part to non-linear lighting distribution across the sandbox, which could be corrected by orientating the backlights effectively. Time savings of around 35% were achieved for this experimental study and this is expected to increase for larger scale studies. The new calibration approach exhibits some promising features in terms of its robustness to experimental error and its ability to process efficiently large-scale experiments in a shorter time frame.

Effects of mixed physical barrier on residual saltwater removal and groundwater discharge in coastal aquifers

“…The high concentration saltwater captured in the upstream aquifer after the construction of the physical barriers is defined as residual saltwater (Luyun et al, 2009). Experiments and numerical simulations were conducted to investigate the effects of physical barriers on SWI (Abdoulhalik & Ahmed, 2017a;Robinson et al, 2018;Senthilkumar & Elango, 2011). The removal of residual saltwater trapped in an inland aquifer is an important aspect of SWI control; this process involves aquifer cleaning and groundwater reuse and is becoming a severe issue restricting fresh groundwater exploitation in coastal regions.…”

New method to calculate the dynamic factor–flow velocity in Geomorphologic instantaneous unit hydrograph

“…Thus, the model factors should be selected comprehensively. The random forest (RF) algorithm is a suitable for handling the selection and classification work and is commonly used in different fields 37–39 . In this study, the RF model is built to screen out the core factors.…”

Optimizing Laboratory Investigations of Saline Intrusion by Incorporating Machine Learning Techniques

“…A traditional MLT approach would feed the raw data from the eight calibration images (with saltwater concentrations of 0%, 5%, 10%, 20%, 30%, 50%, 70%, and 100%), of one homogeneous aquifer for each utilized bead size, into a sufficiently strong learner that would automatically establish a relationship between LI and SW% [54]. Nevertheless, it is well documented [31,33] that experimental data of saltwater intrusion generated in sandbox setups are characterized by a high percentage of irregularity, mostly due to the deviating light absorbing attributes of the individual beads.…”

“…In the field of groundwater saline intrusion, their implementation has been mostly focused on the prediction of the amount of intruding saltwater [49,50] as well as the efficient management of vulnerable aquifers [51][52][53]. Robinson et al [54] utilised the Random Forests ensemble learning method to shorten the experimental procedure in sandbox investigations. The resulting methodology, while taking less time, generated saltwater concentration fields that significantly deviated from those derived by the traditional pixel wise process.…”