Different inputs and ensembles#

Compare simulation cases, calculate differences between models, and summarize ensembles of related runs.

Prepare the comparison cases#

This example uses two SPE11B simulations with different injection rates. The base case is written to spe11b and the modified case to spe11b_larger_inj. The maintained Bash script prepares and runs both cases before generating the figures below.

Compare summary quantities#

Pass several input cases to plopm -i and separate plotted variables with commas. The following command compares three related quantities:

plopm -i 'spe11b/SPE11B spe11b_larger_inj/SPE11B' -v 'fgmip,fgmip / 1E6,RGMDS:5' -yl '[kg]  [Kt]  [kg]' -tu w -fs 10,5 -c r,b -ls 'solid,dashed' -t 'Field gas mass in place  Converted to kilotonnes  Dissolved CO$_2$ in facies 5' -fz 14 -sg 2,2 -rdl 1 -ll empty,empty,empty,center -fn comparison
Summary quantities from two SPE11B simulations

Field gas mass in place, the same quantity converted to kilotonnes, and dissolved CO2 mass in facies 5 for the two simulations.#

The command demonstrates that variable expressions can be entered directly in plopm -v. Here, fgmip / 1E6 converts field gas mass in place from kilograms to kilotonnes.

The remaining options control the comparison:

  • plopm -yl assigns a y-axis label to each quantity.

  • plopm -tu displays time in weeks.

  • plopm -c and plopm -ls distinguish the two cases by color and line style.

  • plopm -sg arranges the three plots in a two-by-two layout.

  • plopm -rdl removes repeated subplot labels.

  • plopm -ll removes legends from the first three panels and places the shared legend in the remaining panel.

Subtract one case from another#

Use plopm -di to subtract a reference case from the primary input. This command calculates the difference in gas saturation at restart step 3:

plopm -i spe11b_larger_inj/SPE11B -v sgas -r 3 -di spe11b/SPE11B -hide 0,0,0,1
Difference in gas saturation between two SPE11B simulations

Gas saturation in the higher-injection case minus gas saturation in the base case.#

The input selected with plopm -di must provide the same variable and a compatible grid. Positive values indicate a larger gas saturation in the primary input; negative values indicate a larger value in the reference case.

Format the difference map#

Customize the difference map with a colormap, fixed color limits, colorbar tick count, number format, and output filename:

plopm -i spe11b_larger_inj/SPE11B -v sgas -r 3 -di spe11b/SPE11B -hide 0,0,0,1 -c tab20c_r -cl '[0,0.8]' -cbn 9 -cbf 0.1 -fn formated
Formatted difference in gas saturation between two SPE11B simulations

Formatted gas-saturation difference with a fixed color range and nine colorbar ticks.#

Use plopm -cl when several difference maps must use the same color range. This makes the magnitudes directly comparable between figures.

Plot ensemble statistics#

Use plopm -ens to summarize a collection of simulations. Ensemble members are identified from the supplied input folders, and the selected mode controls which statistical representation is plotted:

  • -ens 0 disables ensemble processing.

  • -ens 1 plots the ensemble mean with error bands.

  • -ens 2 plots the minimum, mean, and maximum.

  • -ens 3 combines both representations.

Use plopm -fb with modes 1 and 3 to customize the fill colors and alpha values of the error bands.

Statistical summary of a simulation ensemble

Ensemble statistics generated from multiple related simulations.#

The ensemble example directory contains the supporting configuration and generation workflow. The maintained script also demonstrates the available plopm -ens modes.

Reproduce this example#

Run the complete workflow from the repository root:

. ./tests/scripts/docs_different_files_and_ensembles.sh

View script

View raw script

Back to the examples gallery