Oceans: Abstract Values vs. Measured Values - 5

Fig. 1 Abstract maximum, average, and minimum

Just in case the message is not getting through, today I present some more graphs to show how true the three scientific papers (which I quoted from yesterday) were  (Oceans: Abstract Values vs. Measured Values - 4).

I mean where those papers pointed out how the measurements which scientists have been able to take are not spread out evenly (in terms of latitude & longitude) across the vast oceans of the world.

My argument or discussion about this is that we need to have a way of doing with WOD datasets what the GISTEMP and PSMSL data users have been able to do with those datasets.

Fig. 2a

Fig. 2b

Fig. 2c

That is, to define which WOD layers and/or zones can be used to represent the entirety of the ocean conditions, that is, the oceans as a whole.

Only twenty-three PSMSL tide gauge stations out of about 1,400 tide gauge stations can do that in terms of sea level change.

The GISTEMP is similar in that the global mean average temperature anomaly can be shown in the same manner (a representative subset).

But, as those papers discussing the world oceans point out, the same is not yet accomplished with WOD datasets.

The measurements are too concentrated in certain areas to the exclusion of other areas, are too few, and do not go deep enough into the abyss.

The ARGO automated system of submarine drones is changing that in the upper 2,000 m of the oceans, but that is a relatively recent technological win.

There is no long term in situ set of measurements of ocean temperature and salinity data going back in time for over a century, like there are with the PSMSL and GISTEMP datasets.

To visually point out the measurement aberrations I am speaking of, let's look at the new graphs generated by version 1.7 of the software I am constructing.

The graph at Fig. 1 shows the ABSTRACT (calculated) maximum, average, and minimum thermal expansion and contraction pattern from the years 1880 to 2016.

The three graphs at Fig. 2a - Fig. 2c show what happens when in situ WOD measurements are added to the data stream used to generate those abstract graphs.

Fig. 3 Abstract avg. compared to measured

The patterns made by the in situ measurements are out of sync with the abstract patterns made with the WOD information about valid maximum and minimum temperature and salinity values.

Since those values from the WOD manual define validity at all ocean basins and all depths at those basins, being out of sync with them is a problem, especially when the out-of-sync pattern emerges using any of the three different sets of WOD data which compose three different layer lists.

Those three sets are 1) all layers, 2) 6 selected layers, and 3) 8 selected layers, as shown by the report below.

The software module loading sequence proceeds from 1 through 6 (GISS data loader, ABSTRACT data generator, G6 loader, PSMSL loader, G8 loader, and the WOD all-layers loader.

Those modules load in situ measurement data from SQL tables, as well as WOD maximum / minimum valid values.

The software then organizes the data into annual structures (past to present).

The measured values are converted into TEOS values, according to the TEOS rules, by functions in the TEOS toolkit (e.g. Golden 23 Zones Meet TEOS-10).

I may have to stop using WOD layers, to instead use individually selected WOD Zones.

I want to find locations that stay within the guard rails of the valid WOD maximum / minimum values in Appendix 11 of their user manual (see links here).

I am on the case.

The next post in this series is here, the previous post in this series is here.

---

A printout of the loading sequence of the module follows:

---

DREDD BLOG
GISS, PSMSL, WOD & TEOS
Data Analyzer Report
(ver. 1.7)
=======================

(1) GISS Loader
---------------
processed 137 rows


(2) ABSTRACT Calculator
-----------------------
processed:
137 years of data
30 ocean basins
at 33 depths


(3) WOD G6 Loader
-----------------
processing layer 5
processed 118 rows
(59 years) of data

processing layer 7
processed 102 rows
(51 years) of data

processing layer 8
processed 162 rows
(81 years) of data

processing layer 9
processed 176 rows
(88 years) of data

processing layer 10
processed 172 rows
(86 years) of data

processing layer 12
processed 142 rows
(71 years) of data


(4) PSMSL Loader
----------------
processed 10,199 rows


(5) WOD G8 ALT Loader
---------------------
processing layer 3
processed 142 rows
(71 years) of data

processing layer 5
processed 118 rows
(59 years) of data

processing layer 7
processed 102 rows
(51 years) of data

processing layer 8
processed 162 rows
(81 years) of data

processing layer 9
processed 176 rows
(88 years) of data

processing layer 10
processed 172 rows
(86 years) of data

processing layer 12
processed 142 rows
(71 years) of data

processing layer 14
processed 142 rows
(71 years) of data


(6) WOD Loader (all layers)
---------------------------
processing layer 0
processed 60 rows
(30 years) of data

processing layer 1
processed 118 rows
(59 years) of data

processing layer 2
processed 96 rows
(48 years) of data

processing layer 3
processed 142 rows
(71 years) of data

processing layer 4
processed 188 rows
(94 years) of data

processing layer 5
processed 118 rows
(59 years) of data

processing layer 6
processed 178 rows
(89 years) of data

processing layer 7
processed 102 rows
(51 years) of data

processing layer 8
processed 162 rows
(81 years) of data

processing layer 9
processed 176 rows
(88 years) of data

processing layer 10
processed 172 rows
(86 years) of data

processing layer 11
processed 142 rows
(71 years) of data

processing layer 12
processed 142 rows
(71 years) of data

processing layer 13
processed 144 rows
(72 years) of data

processing layer 14
processed 142 rows
(71 years) of data

processing layer 15
processed 156 rows
(78 years) of data

processing layer 16
processed 102 rows
(51 years) of data

processing layer 17
processed 0 rows
(0 years) of data