Skip to main content
Fig. 13 | Advanced Modeling and Simulation in Engineering Sciences

Fig. 13

From: Solving multiphysics-based inverse problems with learned surrogates and constraints

Fig. 13

CO2 plume estimation and forecast using FNO surrogates and NF constraints to invert different modalities of observed data. The first three columns represent past CO2 saturations at day 400, 500, and 600 of the first 600 days of CO2 saturation monitored either through the well measurements or time-lapse data. The last two columns include forecasts for the saturations at future days 700 and 800, where no observed data is available. The first row shows the past and future CO2 estimates yielded by inverting well measurements only. The second row is the same but now inverting time-lapse seismic data. The third row is the same but now jointly inverting well measurements and time-lapse seismic data. The fourth, fifth, and sixth rows show \(5\times \) difference between the ground truth CO2 plume (first row of Fig. 7) and the first, second, third row, respectively. The S/Ns for the first, the second, and the third rows are 15.26 dB, 20.14 dB, 20.46 dB, respectively

Back to article page