mismatch between model and validation plots generated by `plot_selected_validations()`

[FariborzDaneshvar-NOAA]

Here are example of outputs I got for 19 ensembles of Florence 2018, BEST track with inland hydrology	GAHM + NWM

Interestingly, results for the model set up with GAHM and Holland look fine (see below), but it was not the case for GAHM+NWM and GAHM+NWM+WWM scenarios:	GAHM

Path to four scenarios:

• Holland: /lustre/hurricanes/florence_2018_BEST_agu_holland/setup/ensemble.dir/analyze/step_linear_k1_p1_n0.025
• GAHM: /lustre/hurricanes/florence_2018_BEST_agu_gahm/setup/ensemble.dir/analyze/step_linear_k1_p1_n0.025
• GAHM+NWM: /lustre/hurricanes/florence_2018_BEST_agu_gahm_nwm/setup/ensemble.dir/analyze/step_linear_k1_p1_n0.025
• GAHM+NWM+WWM: /lustre/hurricanes/florence_2018_BEST_agu_gahm_nwm_wwm/setup/ensemble.dir/analyze/step_linear_k1_p1_n0.025

[FariborzDaneshvar-NOAA]

Another example of a similar mismatch with 10 ensemble without inland hydrology! https://github.com/noaa-ocs-modeling/EnsemblePerturbation/issues/104#issuecomment-1652362777