Three trajectories on the same grids and times.
After placement (§01) and the per-window 3D-Var cycling (§02), the pipeline
releases a free forecast out to 21 h driven by ndown from
d01 → d02 → d03 → d04. The reflectivity truth is the INEA Macaé composite —
the same radar whose volumes were assimilated (not an independent truth);
rain gauges (INMET / Alerta Rio / CEMADEN) provide the independent
precipitation check.
Per-domain plots (d01, d02, d03, d04) · cycled vs control vs radar truth · CSI / POD / FAR / bias heatmaps and time series
Root: containers/_instances/assimilator/fcst_cycled_2024012100/
cycle_00z/, cycle_01z/,
cycle_02z/, cycle_03z/cycle_06z/, cycle_07z/,
cycle_08z/, cycle_09z/Per cycle: wrfvar_output (analysis), fg
(background), wrfbdy_d01, statistics,
cost_fn, radar_omb_oma_01.
Same root:
free_fcst_radar/ — morning radar (03 Z → 00 Z + 1)
readyfree_fcst_radar_afternoon/ — afternoon radar
(09 Z → 06 Z + 1) readyfree_fcst_control_afternoon/ — afternoon control
readyfree_fcst_control does not exist as a dedicated d01
dir. For the 1:1 morning d01 control comparison, use
/experiments/RADAR_DA_FCST_4DOM_2024012100/runs/real_4dom_morning/
(same GFS, no DA).Root: containers/_instances/assimilator/fcst_ndown_2024012100/
radar/d02, radar/d03,
radar/d04 ready —
control/d02, control/d03, control/d04
ready.radar_afternoon/d02, radar_afternoon/d03,
radar_afternoon/d04 ready —
control_afternoon/d02, control_afternoon/d03,
control_afternoon/d04 ready.Each dX/ contains wrfout_d01_2024-01-21_09:00:00 ..
2024-01-22_06:00:00 (afternoon) or the 03 Z → 00 Z + 1 analogue
(morning). Note: after ndown, the child was renamed to
d01 in the namelist.
control_afternoon/d04/wrfout_d01_2024-01-22_06:00:00 on disk).
The full d01 / d02 / d03 / d04 control vs radar comparison is unblocked.
docs/RADAR_DA_MACAE_2024012100_PIPELINE.md
in the CONVECT repo — covers grids, dates, and ndown gotchas.containers/plotter/scripts/wrfda_verification/make_da_innovation_report.py