Table Of Contents

This Page

Net Lagrangian PV change in exp6 with 1/10-day outputΒΆ

We perform the same analysis than in this note for exp6 using 1/10-day model output. The error due to the analysis is reduced as the estimates from snapshots and averages come closer (Fig. 1), compared to previous calculation with model outputs of longer temporal resolution. However, what is corrected is the estimate from snapshots as the error estimated from averages does not change (compare Fig. 1b here with Fig. 4b of this note). This suggests that the error estimated from averages is due mainly to the error in the model and cannot be reduced. It is, thus, still puzzling that the error in exp4, a simulation that does not satisfy well the PV balance during a cycle, has the same magnitude than the error in exp6.


Figure 1: (a) Net Lagrangian PV change for 4 parcels in exp6 using the 1/10-day model output (blue line) from Q computed with snapshots u and v, (blue dots) from Q computed with averaged output Q, (red line) from HOR computed with snapshots u and v, (red dots) from HOR computed with averaged output HOR. (b) Error in the net PV change. Each color corresponds to a parcel: (lines) using u and v snapshots, (dots) using averages. See analysis_2_script.m in RESEARCH/MODELISATION/HIM/studies/PV_and_dissipation/forced_damped_wave/exp4/quarter_of_a_day_output.