Glitch in monthly mean diagnostics

[penguian]

| by mrd599@nci.org.au

---

Arnold noticed an odd pattern in the standard deviation of mslp in Roger's u-as986 AMIP run.

It looks like this is due to a glitch in the monthly mean diagnostics on one processor.

We need to be sure our QC procedures can catch this.

---

Issue migrated from trac:358 at 2024-01-31 18:33:40 +1100

[penguian]

@martin.dix@anu.edu.au uploaded file as896.psl.std.dev.gif (33.8 KiB)

[penguian]

@martin.dix@anu.edu.au _uploaded file ts_series.png (65.2 KiB)_

[penguian]

@martin.dix@anu.edu.au _uploaded file ts_map.png (60.4 KiB)_

[penguian]

@martin.dix@anu.edu.au _uploaded file ts_map_daily.png (69.5 KiB)_

[penguian]

@martin.dix@anu.edu.au commented

---

Arnold's original map of SD of mslp

Time series of monthly mean surface temperature from a point in the problem region shows the problem occurs in a single month.

Map of monthly mean surface temperature for April 1981. This region is of the right size to be a single processor.

Map of monthly mean surface temperature for April 1981, calculated from the daily means

All fields in the Apr 1981 monthly mean file show the problem. Values in the region are ~ 1/15 less than the surrounding region, suggesting that the monthly mean calculation on one processor has somehow dropped two days from the month.

[penguian]

@martin.dix@anu.edu.au edited the issue description

[penguian]

@martin.dix@anu.edu.au commented

---

In u-av842, the monthly mean file for one month was incomplete aand only about 1/3 the size it should have been. Other files for the month were ok.