Table Of Contents

This Page

03.24.13: LPV analysis of exp2_l2ΒΆ

Here, I show the results of various LPV analysis for exp2_l2. Figs. 1 and 2 show the LPV analysis for two different periods (days 2515-2615 and days 8015-8115, respectively). As we show in 03.22.13: Long term behavior of exp2_l2, the solution has not yet reached a stationary steady state over days 2515-2615 and that shows up with large dQL/dt compared to the period days 8015-8115.

Fig. 3 shows that the calculation of the advective term is sensitive to the horizontal resolution. This makes sense if we think that with weak wave amplitude, the displacement of parcels is small and the calculation becomes even more sensitive to the interpolation scheme.

Finally, although Fig. 4 shows that the analysis is not too dependent on the choice of the output (1-day averages in Fig. 4 versus 1-day output in Fig. 2), Fig. 5 shows that the LPV analysis may nonetheless be sensitive to the temporal resolution as well.

Note that, except in the last case (Fig. 5), the dissipative term stays insensitive to the temporal and spatial resolution. This suggests that, even if we do not get the balance, we could use only the dissipative term as a way to infer the circulation. Getting the balance, however, is important to get in order to demonstrate that the model does satisfy the PV balance.


Figure 1: LPV analysis for days 2515-2615.


Figure 2: LPV analysis for days 8015-8115.


Figure 3: LPV analysis for days 8015-8115 using every 2 points in the horizontal space.


Figure 4: LPV analysis for days 8015-8115 using 1-day averaged outputs.


Figure 5: LPV analysis for days 8015-8115 using every 2-day outputs.

These figures were produced with plot_LPV_analysis.m using Matlab files produced by theory_test_several_cycles_script.m both in RESEARCH/MODELISATION/HIM/studies/diss_train_of_eddies/exp2/exp2_l2/analysis_1d on the main disk on ipu1.